Вы находитесь на странице: 1из 18

International Journal of Hospitality Management 50 (2015) 9–26

Journal of Hospitality Management 50 (2015) 9–26 Contents lists available at ScienceDirect International

Contents lists available at ScienceDirect

International Journal of Hospitality Management

journal homepage: www.elsevier.com/locate/ijhosman

journal homepage: www.elsevier.com/locate/ijhosman Why is hospitality employees’ psychological capital

Why is hospitality employees’ psychological capital important? The effects of psychological capital on work engagement and employee morale

Soyon Paek a,1 , Markus Schuckert b,2 , Taegoo Terry Kim c, , Gyehee Lee c,3

a Faculty of Hospitality and Tourism Management, Macau University of Science and Technology, Avendia Wai Long, Taipa, Macau

b School of Hotel and Tourism Management, The Hong Kong Polytechnic University, 17 Science Museum Road, TST-East, Kowloon, Hong Kong SAR, China

c Department of Tourism Management, College of Hotel and Tourism Management, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 130-701, Republic of Korea

Dongdaemun-gu, Seoul 130-701, Republic of Korea article info Article history: Received 24 February 2014

article

info

Article history:

Received 24 February 2014 Received in revised form 29 April 2015 Accepted 6 July 2015

Keywords:

Psychological capital Personal resources Work engagement Employee morale Conservation of resources theory Job demands-resources model

abstract

This study examines work engagement as a partial mediator of the effect of psychological capital (Psy- Cap) on employee morale in a sample of hotel employees. A survey was carried out with 312 front-line staff from 15 five-star hotels in Seoul, Korea. A one-month time-lag design (Time 1: PsyCap and work engagement; Time 2: employee morale) was used to reduce potential common method bias. The hypoth- esized relationships in the model were tested using structural equation modeling. The results suggest that work engagement partially mediates the effect of PsyCap on job satisfaction and affective organi- zational commitment. Specifically, front-line employees with high PsyCap are more engaged with their work and more likely to display job satisfaction and affective organizational commitment. The study concludes with a discussion of its empirical findings, strengths, theoretical contributions, and practical implications. Limitations and their implications for future studies are also reviewed. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Why is psychological capital (PsyCap) relevant? As Maltz et al., 2003 point out, measuring organizational success is a continuous challenge, requiring the use of different assessment approaches which take both monetary and nonmonetary criteria into account. Measuring various types of capital is one such approach. In other words, the concept of capital is no longer confined to the mone- tary/financial context (previously the dominant approach), but now extends towards a more differentiated understanding (Anheier et al., 1995). Based on this development, it has been argued that in order to be successful and sustainable, an enterprise needs to assess different forms of capital, such as human, cultural, social, or reputational. However, according to scholars such as Luthans et al. (2008) and Avey et al. (2009), these approaches fail to measure

Corresponding author. Fax: +82 2 964 2537. E-mail addresses: spaek@must.edu.mo (S. Paek), markus.schuckert@polyu.edu.hk (M. Schuckert), tgkim@khu.ac.kr (T.T. Kim), ghlee@khu.ac.kr (G. Lee).

1 Fax: +853 2882 5990.

2 Fax: +852 2362 9362.

3 Fax: +82 2 964 2537.

http://dx.doi.org/10.1016/j.ijhm.2015.07.001

0278-4319/© 2015 Elsevier Ltd. All rights reserved.

another critical aspect, namely the psychological and attitudinal strengths of individuals as identified through components such as motivation, work engagement, job satisfaction, and affective orga- nizational commitment. Since existing paradigms do not properly incorporate these employee assets, Luthans and Youssef (2007) and Nelson and Cooper (2007) have driven the development of PsyCap as a construct. They come from the perspective of positive organiza- tional behavior, which in turn is based on the positive psychology of Peterson and Seligman (2004). In this context, the concept of Psy- Cap designates and measures the different behavioral states that are ultimately relevant to the performance of an employee within an organization (Luthans et al., 2007b). Excellent employee performance is important and desirable in all industries and sectors, but in the labor-intensive service indus- tries, employees are a particularly important part of the product and form the core of the service experience (Slåtten and Mehmetoglu, 2011). Hospitality staff can deliver competitive advantage in terms of building and maintaining host-guest relationships (Onsøyen et al., 2009) and quality, and building guest loyalty (Chi and Gursoy, 2009). Highly motivated and engaged employees are critical to the success of service organizations and enterprises (Bakker and Demerouti, 2008; Slåtten and Mehmetoglu, 2011).

10

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Self-efficacy

Optimism

Hope

Resilience

Job Satisfaction

Time 2

Hypothesis 2 (+)

Partial Mediator Hypotheses 6 and 7 PsyCap Work Engagement Hypothesis 1 (+) Time 1 Time
Partial Mediator
Hypotheses 6 and 7
PsyCap
Work Engagement
Hypothesis 1 (+)
Time 1
Time 1

Hypothesis 4 (+)

Hypothesis 5 (+)

Hypothesis 3 (+)

Affective Organizational

Commitment

Time 2

Fig. 1. Research model and hypotheses of work engagement as a partial mediator of the relationship between PsyCap and employee morale.

The virtuous circle of employee performance, perceived qual- ity, profit, and management support is well characterized in the service management literature (Grönroos, 2000). Based on this chain of cause and effect, recent research in the hospitality context places the employee and his/her settings, motivations, engage- ment, and satisfaction with the work environment and conditions at the center stage, not only conceptually but empirically (Slåtten and Mehmetoglu, 2011). In the hospitality industry in particular, employees’ mental outlook, mood, and behavior are very important as they exert a critical influence on performance, outcomes, and customer satisfaction. Such engagement affects the service climate and employees’ achievements as well as customer loyalty (Salanova et al., 2005). Among the several attitudinal and behavioral factors relevant to this, the concept of work engagement is particularly prominent, given its strong relationship with work performance and competitive advantage (Baumruk, 2004). Bakker and Demerouti’s (2007, 2008) job demands-resources (JD-R) model proposes that job resources such as personal psycho- logical resources lead to improved employee engagement and in turn positively affect job-related outcomes. Since then, in the man- agement literature, meaningful initial academic attention has been paid to the investigation of the relationships between personal psy- chological factors, work engagement, and job-related outcomes. In the hospitality domain, despite the theoretical and manage- rial significance of the link of personal psychological factors, work engagement, and job-related outcomes, as employees’ competitive advantage is pivotal in the success of hospitality firms, there is little empirical evidence of the holistic relationships among these vari- ables. Only a few recent studies in the hospitality literature have examined a part of these relationships (e.g., Karatepe and Olugbade, 2009; Karatepe et al., 2006) or dealt with limited sub-constructs of personal psychological factors, work engagement, and job-related outcomes (e.g., Karatepe, 2014). These studies tested hope, trait competitiveness, and/or self-efficacy. Therefore, the originality and significance of the present study lies in its holistic examination of personality-related constructs, work engagement, and job-related outcomes. It does this by aligning work engagement with PsyCap as an antecedent and measuring as outcomes the effect on job sat- isfaction and affective organizational commitment, as components of employee morale. Furthermore, as will be discussed below, the role of work engagement as a determinant of job-related outcomes has still not been fully researched or understood (Christian et al., 2011; Karatepe et al., 2013). In the hospitality context, empirical evidence of its antecedents and consequences is also lacking (Karatepe, 2011; Slåtten and Mehmetoglu, 2011). In addition, among personality- related constructs, PsyCap has been considered separately from

work engagement in most past empirical studies. While both are considered highly relevant to, and important in, achieving posi- tive work outcomes, such separation leads to the idea that they are limited. Therefore, we propose that there is a full and signifi- cant link between PsyCap, work engagement, and outcomes such as employee morale. In extending the research base, this study therefore attempts to close these gaps by investigating the partial mediating role of work engagement in the relationship between PsyCap and employee morale as represented by job satisfaction and affective organiza- tional commitment in a hospitality environment, based on data collected from front-line employees in top-tier hotels. It draws on the conservation of resources (COR) theory (Hobfoll, 1989, 2001) and the JD-R model of work engagement (Bakker and Demerouti, 2007, 2008) to examine the following research questions:

Research question 1: Does PsyCap directly and positively pre- dict work engagement and employee morale?

Research question 2: Does work engagement, in turn, directly and positively predict employee morale?

Research question 3: Does work engagement partially mediate the effect of PsyCap on employee morale?

2. Theoretical foundation, research model, and hypotheses

Fig. 1 presents the research model setting out the hypothesized relationships. It is proposed that employees’ PsyCap is positively related to their work engagement (Hypothesis 1). Furthermore, work engagement acts as a central “switch” and predictor of employee morale (i.e., job satisfaction and affective organizational commitment; Hypotheses 4 and 5), such that there will be a positive relationship between PsyCap and employee morale as mediated by work engagement (Hypotheses 6 and 7). However, the model also proposes that work engagement is only a partial mediator, because there is a direct relationship between PsyCap and the two com- ponents of employee morale (Hypotheses 2 and 3). At a glance, this proposed model is clearly innovative in terms of its analysis of employee morale-related job outcomes based on PsyCap, partially mediated through work engagement as a core concept of the JD-R model.

2.1. The COR theory and the JD-R model as underlying

frameworks

In order to outline the proposed approach, we explain the role of PsyCap as an antecedent of work engagement using the COR theory

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

11

Innovative Stress symptoms, job search behavior, and intentions to quit (Avey et al., performance (Abbas
Innovative
Stress
symptoms, job
search behavior,
and intentions
to quit
(Avey et al.,
performance
(Abbas and Raja,
Employee well-
2011)
being (Avey et
al., 2010)
2009)
Job performance
and quality of
work life
(Nguyen and
Nguyen, 2012)
Job satisfaction
and performance
(Luthans et al.,
2007a)
Organizational
Innovative
trust
behavior (Jafri,
(Walumbwa et
2012)
al., 2011)
Commitment to
Trust in
organizational
management
mission (Luthans
and performance
and Jensen,
(Clapp-Smith et
2005)
al., 2009)
PsyCap
Job satisfaction
Burnout and
and
organizational
emotional labor
commitment
(Cheung et al.,
(Larson and
2011)
Luthans, 2006)
Supportive
organizational
climate for
employees
(Luthans et al.,
Perceived trust
(Norman et al.,
2010)
2008b)
Performance and
Workplace
organizational
performance
citizenship
(Luthans et al.,
behaviors (Gooty
Turnover and
et al., 2009).
2008)
absence
Creativity
intentions
(Rego et al.,
(Karatepe and
2012)
Karadas, 2014)

Fig. 2. PsyCap sun: previous empirical research related to PsyCap.

and the JD-R model as theoretical frameworks. The central tenet of the COR theory is that individuals strive to obtain, maintain, and preserve certain resources which they value (Hobfoll, 1989, 2001). Such resources can be defined as “those objects, personal character- istics, conditions, or energies that are valued by the individuals or that serve as a means for attainment of these objects, personal char- acteristics, conditions, or energies” (Hobfoll, 1989, p. 516). Just as with material goods, individuals also seek to acquire, maintain, pre- serve, and accumulate immaterial or intangible resources (Bakker and Leiter, 2010; Salanova et al., 2005). Hobfoll (2001) explains that the gain and loss of these resources is not symmetric, as “a loss is disproportionately more salient than resource gain” (p. 343). These resources can be depleted and they must be invested in to pro- tect against or recover from a loss or to make a gain. Individuals with more resources have less fear of loss and are more capable of managing gain (Hobfoll, 2001). Hobfoll (2001) also proposes that resources tend to generate each other since one may possess one major type of resource that is linked with, or can replace, others. For example, when a role is demanding, job resources like social support are linked to, or may even replace, personal resources such as self-efficacy or optimism. Hobfoll calls this linkage and interplay “resource caravans” (2001, p. 349). Acquiring increasing amounts thus creates such resource caravans, leading to positive outcomes (Hobfoll, 2001). This notion is connected to the JD-R model proposed by Bakker and Demerouti (2007, 2008). This divides job characteristics into two general cate-

gories: job-related demands and resources (Bakker and Demerouti, 2007, 2008; Bakker and Leiter, 2010). Job-related demands are the physical, psychological, social, or organizational aspects of a job, which require sustained physical, cognitive, and emotional effort or skills. They include work pressure and emotional, men- tal, or physical demands. Job-related resources can be separated into the physical, psychological, social, or organizational aspects of job autonomy, performance feedback, or social support; and per- sonal resources, which are a psychological state of development related to optimism, hope, or resilience (Bakker and Demerouti, 2007, 2008). These personal resources are known as PsyCap. The JD-R model suggests that job resources play a salient role in initiating a motivational process that leads to employee work engagement and in turn enhances job performance (Bakker and Demerouti, 2007, 2008). Job resources can function as intrinsic motivators and extrinsic motivators. Job resources as intrinsic motivators, such as colleague or supervisor support, satisfy basic human needs and foster employees’ growth and development. Job resources as extrinsic motivators promote employees well-being in order to boost their efforts to fulfill their job requirements. The model emphasizes the creation of “resource caravans” when individuals with abundant job resources tend to have a more positive psychology. Such individuals feel higher level of concor- dance between their goals and themselves. Consequently, these employees are intrinsically and/or extrinsically motivated to con- centrate on their goals and drive themselves to engage with

12

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

their work to a higher level (Xanthopoulou et al., 2009). In turn, higher work engagement generates positive organizational out- comes. The motivational process of the JD-R model indicates the role of work engagement as a mediator in the relationship between job resources and organizational outcomes such as organizational commitment (Llorens et al., 2007).

2.2. Relationship between PsyCap and work engagement

Sweetman and Luthans (2010) suggest that with the increase in popularity of evidence-based management, understanding the process of employee engagement has become critical for mod- ern organizations. A number of researchers, primarily Luthans and Youssef (2007), developed the construct of PsyCap as an out- growth of positive organizational behavior and have shown that the psychological capacities of human resources can be mea- sured, developed, and efficiently managed. This approach has its roots in the theory of positive psychology proposed by Peterson and Seligman (2004). PsyCap has been conceptually identified by Luthans and colleagues (Luthans and Jensen, 2005; Luthans et al., 2007a,b) as consisting of the four positive psychological resources of self-efficacy, optimism, hope, and resilience. Luthans et al. (2007b) proposed that the second-order factor of PsyCap may represent the common source of variance (i.e., common mechanis- tic processes) connecting the four constructs (i.e., hope, optimism, resilience, and self-efficacy). The comprehensive definition of Psy- Cap is “an individual’s positive psychological state of development characterized by: (1) having confidence (self-efficacy) to take on and put in the necessary effort to succeed at challenging tasks; (2) making a positive attribution (optimism) about succeeding now and in the future; (3) persevering toward goals and, when neces- sary, redirecting paths to goals (hope) in order to succeed; and (4) when beset by problems and adversity, sustaining and bouncing back and even beyond (resilience) to attain success” (Luthans et al., 2007b, p. 3). Luthans et al. (2005, 2007b) have proved that PsyCap predicts related outcomes more accurately than its stand-alone compo- nents. This has been confirmed by several studies which have tested various component-outcome combinations (Luthans et al., 2005, 2007a; Stajkovic and Luthans 1998). Since its introduction and con- ceptual formulation, empirical studies have analyzed the role of PsyCap as an important predictor or mediator of various work- related outcomes. As summarized in Fig. 2, PsyCap is related to behavioral and psychological factors as well as to entrepreneurial, managerial, and economic outcomes. In all of these studies, PsyCap has been shown either to have a significant positive relationship to outcomes or to reduce negative effects. In addition, our analysis of the related literature shows that none of these empirical studies of antecedents and outcomes has considered work engagement. Work engagement was first conceptualized by Kahn (1990) and then operationalized by Maslach and Leiter (1997). Schaufeli et al. (2002) and Schaufeli and Bakker (2004) then adjusted the concept, which is now characterized as “a positive, fulfilling, work-related state of mind” (Schaufeli and Bakker, 2004, p. 295). The engaged person is enthusiastic, inspired, proud, and challenged at work, and is willing to make an effort while maintaining concentration and being deeply engrossed in the task. He or she is also persistent in the face of difficulties and distractions, such as time at work passing quickly, and it can be difficult for him or her to detach him or herself from a task (Schaufeli et al., 2002, pp. 74–75). According to Kahn (1990), work engagement shapes the process of how staff choose to be present and absent during task performance. Sweetman and Luthans (2010) present a conceptual model which relates PsyCap to work engagement through positive emo- tions, which are part of the JD-R model (Bakker and Demerouti, 2007; Schaufeli and Bakker, 2004). They argue that PsyCap could

have a direct and state-like relationship to work engagement, and should not be considered as a fluctuation or a fleeting shift such as a mood or emotion (e.g., anger or joy). Recent research in the hospitality industry suggests that work engagement may have dif- ferent antecedents (Table 1): for instance, Slåtten and Mehmetoglu (2011) use role benefit, job autonomy, and strategic attention, while Li et al. (2012) look at the exchange between managers and staff and the consistency of human resource management. In addi- tion, Karatepe links work engagement to high-performance work practices (Karatepe, 2013a) or perceptions of organizational poli- tics (Karatepe, 2013b). Motivated by this body of research on the antecedents of work engagement in the hospitality industry, and with the theoretical basis of the JD-R model, we propose that Psy- Cap is an important driver of, and directly influences, work-related outcomes. Based on the view of PsyCap as a personal resource, as identified by COR theory (Hobfoll, 1989, 2001; Karatepe and Olugbade, 2009), we propose our first hypothesis as follows:

Hypothesis 1.

PsyCap has a positive influence on work engage-

ment.

2.3. Direct effect of PsyCap on employee morale

As the impact of ethics has become critical in modern busi- ness settings, researchers have undertaken a serious examination of workers’ ethics and morale as outcome variables. Under this umbrella, along with individual morale frameworks, individuals’ attitudes and behaviors in terms of issues such as satisfaction and commitment have been a central topic (Ambrose et al., 2008). Based on this strand of research, by taking major constructs which cap- ture employee morale and are researched in the area of employee morale (i.e., job satisfaction and affective organizational commit- ment) as outcome variables, we attempt here to investigate the relationship between PsyCap and these two constructs, job satis- faction and affective organizational commitment. Job satisfaction

is defined here as “the positive emotional state resulting from the

appraisal of one’s job or job experiences” (Locke, 1976, p. 1300). Affective organizational commitment is defined as “an affective or emotional attachment to the organization such that the strongly committed individual identifies with, is involved in, and enjoys membership in, the organization” (Allen and Meyer, 1990, p. 2). Sweetman and Luthans (2010) identify PsyCap as a resource which generates increasing awareness of, and sensitivity to, employees’ own resources as well as organizational dimensions

and job-related outcomes. Several studies show that PsyCap has

a positive effect on employees’ work attitudes, including job sat-

isfaction (e.g., Larson and Luthans 2006; Luthans et al., 2007a) and organizational commitment (e.g., Larson and Luthans 2006), as well as organization-related dimensions like perceived trust (e.g., Norman et al., 2010), organizational citizenship behavior (e.g., Gooty et al., 2009), and employee perceptions of the organiza- tional climate as supportive (e.g., Luthans et al., 2008). At a glance, therefore, the empirical evidence shows that PsyCap affects indi- vidual satisfaction with work as well as one’s commitment to it. Yet, in a hospitality context, the relationship between PsyCap and employee morale has to be researched specifically. However, based on the discussion presented above of the positive effect of PsyCap on employee attitudes and behaviors in a mainstream management context, we propose that:

Hypothesis 2.

PsyCap has a positive influence on job satisfaction.

Hypothesis 3.

PsyCap has a positive influence on affective orga-

nizational commitment.

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

13

Table 1 Previous empirical studies of the relationship structure of work engagement in the hospitality industry.

Author (year)

Regional

Sample

Relationship structure of work engagement

Main results

context

Salanova et al. (2005)

Spain

Customer contact employees from hotels and restaurants

A: Organizational resources and work engagement M: Service climate

Service climate fully mediates the effects of organizational resources and work engagement on employee performance and customer loyalty.

 

(n

= 342)

C: Employee performance and customer loyalty A: Burnout, engagement, and coping C: General health

Pienaar and Willemse

South

Waiters, waitresses, and bartenders from restaurants and coffee shops (n = 150)

General health of service staff can be predicted by their feelings of personal accomplishment and work engagement (dedication), having avoidant coping strategies, and by favoring the addressing of symptoms in coping. Work engagement (dedication) predicts work (such as job satisfaction and intention to quit) and psychological well-being outcomes. The most prominent traits predicting work engagement are conscientiousness and neuroticism, and the most critical personality trait affecting burnout is neuroticism.

(2008)

Africa

Burke et al. (2009)

China

Hotel managers

A: Work engagement

 

(n

= 309)

C: Work and psychological well-being outcomes A: Big Five personality factors C: Work engagement and burnout

Kim et al. (2009)

USA

Managerial/supervisory and nonsupervisory employees from quick-service restaurants (n = 187) Front-line hotel employees and supervisors (n = 143)

Karatepe (2011)

Nigeria

A: Procedural justice M: Work engagement C: Affective organizational commitment, job performance, and extra-role customer service A: Role benefit, job autonomy, and strategic attention M: Work engagement C: Innovative behavior A: Work engagement M: Job embeddedness C: Organizationally valued job outcomes (turnover intention and job performance) A: Leader-member exchange and human resource management consistency M: Work engagement C: Employee job performance

Work engagement fully mediates the effects of procedural justice on affective organizational commitment, job performance, and extra-role customer service.

Slåtten and

Norway

Front-line hospitality employees (n = 279)

Work engagement is closely and positively linked to employees’ innovative behavior. Role benefit, job autonomy, and strategic attention are positively related to work engagement. Job embeddedness partially mediates the impact of work engagement on turnover intention and job performance.

Mehmetoglu (2011)

Karatepe and Ngeche

Cameroon

Front-line employees from four- and five-star hotels (n = 212)

(2012)

Li et al. (2012)

China

298 employees and 54 supervisors from a luxury hotel (n = 352)

Leader-member exchange is positively related to employee job performance. Moreover, work engagement partially mediates this relationship. Human resource management consistency strengthens the influence of leader-member exchange on work engagement. Levels of work engagement differ depending on the generational membership of employees. The effect of work engagement on turnover intention is moderated by generational differences. Millennials are a more distinct cohort from Gen X-ers and Baby Boomers in terms of their level of work engagement as well as the relationship between work engagement and turnover intention. The effect of polychronicity on job performance and extra-role customer service is fully mediated by work engagement.

Park and Gursoy (2012)

USA and

Customer contact

A: Work engagement M: Generational differences (work values and attitudes)

Canada

employees from mid- and upscale hotels

 

(n

= 677)

C: Turnover intention

Karatepe et al. (2013)

Cyprus

Front-line employees

A: Polychronicity M: Work engagement

 

from five-star hotels

(n

= 185)

C: Job performance and extra-role customer service A: High-performance work practices (appraisal of training, empowerment, and rewards) M: Work engagement C: Job performance and extra-role customer service A: Perceptions of organizational politics M: Work engagement C: Affective organizational commitment, extra-role performance, and turnover intention A: Tourism involvement M: Work engagement C: Job satisfaction

Karatepe (2013a)

Romania

Front-line hotel employees and their managers (n = 110)

Work engagement acts as a full mediator between high-performance work practices and both job performance and extra-role customer service.

Karatepe (2013b)

Iran

Front-line employees and supervisors from four- and five-star hotels (n = 231)

Work engagement acts as a full mediator of the impact of perceptions of organizational politics on affective organizational commitment, extra-role performance, and turnover intention.

Yeh (2013)

Taiwan

Front-line hotel employees (n = 336)

Tourism involvement is positively related to work engagement. Tourism involvement and work engagement are also positively related to job satisfaction. Work engagement partially mediates the relationship between tourism involvement and job satisfaction. Both individual crafting and collaborative crafting are related to job engagement. Person-job fit mediates the relationships.

Chen et al. (2014)

Taiwan

Front-line hotel employees (n = 246)

A: Individual crafting and collaborative crafting M: Person-job fit C: Job engagement

14

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Table 1 (Continued)

Author (year)

Regional

Sample

Relationship structure of work engagement

Main results

context

Karatepe et al. (2014)

Cyprus

Front-line hotel employees (n = 195)

A: Challenge stressors (work overload and job responsibility) M: Work engagement C: Affective organizational commitment and job performance A: Core self-evaluations M: Work engagement C: Work-family facilitation and family-work facilitation

Work engagement fully mediates the effects of challenge stressors on affective organizational commitment and job performance.

Karatepe and Demir

Turkey

Front-line hotel employees (n = 211)

Work engagement fully mediates the effects of core self-evaluation on work-family facilitation and family-work facilitation.

(2014)

Note: A = antecedents; M = mediator; and C = consequences.

2.4. Relationship between work engagement and employee

morale

With a number of recent studies highlighting the importance of work engagement in organizational performance (Leiter and Bakker, 2010), the body of work on this topic (particularly with an international focus) is growing. Recent research has been cross- disciplinary in nature, encompassing the medical sector, hospitality and tourism, financial services, transportation, and retail (Leiter and Bakker, 2010). The work carried out in a hospitality context, in particular, highlights several important aspects of employee morale. The evidence indicates that there is a relationship between staff well-being and service outcomes (Karatepe and Ngeche, 2012; Kim et al., 2009). In addition, hospitality staff are exposed to sub- stantial stress and constant customer demands, which may affect the quality of service delivery (Pienaar and Willemse, 2008). Low employee morale leads to higher operational costs and poorer service performance through lack of adherence to routine and fluc- tuations (Salanova et al., 2005). Finally, a positive staff attitude is critical in facilitating or impeding the implementation of new service initiatives (Karatepe, 2011, 2013a). In empirical terms, a review of the literature on work engage- ment in the hospitality setting shows that it has a positive relationship to several outcomes, such as positive personal con- dition (i.e., general health), and work attitudes, such as job satisfaction (e.g., Burke et al., 2009; Pienaar and Willemse, 2008; Yeh, 2013), as well as organizational achievements such as orga- nizational commitment, job performance, innovative behavior, and extra-role customer service (e.g., Karatepe and Ngeche, 2012; Karatepe, 2013a,b; Karatepe et al., 2013; Li et al., 2012; Slåtten and Mehmetoglu, 2011). It also has a negative relationship to turnover intention (Karatepe and Ngeche, 2012; Karatepe, 2013b; Park and Gursoy, 2012). Therefore, given this body of evidence and building on the previous discussion, we propose that work engage- ment enhances hotel employees’ morale, leading to the following hypotheses:

Hypothesis 4.

Work engagement has a positive influence on job

satisfaction.

Hypothesis 5. Work engagement has a positive influence on affec- tive organizational commitment.

2.5. Partial mediating role of work engagement

We argued above, based on the JD-R model (Bakker and Demerouti, 2007, 2008; Bakker and Leiter, 2010), that PsyCap, as an aspect of job-related personal resources, will positively affect work engagement. Following the JD-R model, Bakker and Leiter (2010) propose that work engagement has a mediating role on per- formance factors such as in- or extra-role performance, creativity, innovation, and financial success. Based on the earlier discussion, we assume that PsyCap will directly predict job satisfaction as well

as affective organizational commitment and work engagement. In addition, given that PsyCap may predict work engagement, which in turn may predict job satisfaction and affective organizational commitment, it is logical to assume that work engagement will act as a partialmediator of the relationship between PsyCap and each of these outcomes. This is consistent with several studies in the fields of management, hospitality, and tourism, in which work engage- ment has been shown to partially mediate the relationship between antecedent and consequent variables (Salanova et al., 2005; Yeh, 2013), although work engagement has been suggested as a full mediator in most studies (Karatepe, 2013a,b, 2014; Karatepe et al., 2013, 2014; Karatepe and Demir, 2014; Leung et al., 2011; Li et al., 2012) or as a mediator in a reciprocal relationship with several variables (Schaufeli and Bakker, 2004; Xanthopoulou et al., 2009). Yeh (2013) examined work engagement as a partial mediator in the relationship between tourism involvement and job satisfac- tion using the data collected from front-line employees in the hotel industry in Taiwan. Salanova and colleagues (2005) linked organi- zational resources and service climate with work engagement as a partial mediator using the data from contact employees in the hotel and restaurant industries in Spain. On the basis of this discussion, we propose the following:

Hypothesis 6.

PsyCap on job satisfaction.

Work engagement partially mediates the effect of

Hypothesis 7.

PsyCap on affective organizational commitment.

Work engagement partially mediates the effect of

3. Methodology

3.1. Common method bias

Common method bias (CMB) refers to the amount of spurious covariance shared among variables due to the common method used in data collection (Buckley et al., 1990). CMB arises from hav- ing a common rater, a common measurement context, a common item context, or from the characteristics of the items themselves (Podsakoff et al., 2003). Therefore, because the data of this study were collected via self-report (i.e., common rater) questionnaires, we need to address certain concerns regarding CMB. If CMB is not controlled in empirical studies, like this one, it is a potential threat to the magnitudes of the hypothesized relationships among the study variables. To minimize and check the potential influence of CMB, we thus used the following four different procedural meth- ods in designing the questionnaires and two statistical remedies in data analysis suggested by Podsakoff and colleagues (Podsakoff and Organ 1986; Podsakoff et al., 2003). First, we assured our study participants of the anonymity and confidentiality of their responses in order to reduce evaluation apprehension and social desirability, which refers to the ten- dency for participants to present a favorable image of themselves (Johnson and Fendrich, 2005). Participants may believe the infor-

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

15

mation they report, or may “fake good” to conform to socially acceptable values, avoid criticism, or gain social approval (King and Bruner, 2000). Therefore, to reduce evaluation apprehension and social desirability, we instructed respondents to complete the

questionnaire individually and then return it to their managers in

a sealed envelope. Second, we collected data from front-line hotel employees with

a time lag of one month (i.e., there were two waves of data col-

lection; Time 1: PsyCap, work engagement, and the demographic profile of respondents; Time 2: job satisfaction and affective orga- nizational commitment) consistent with several empirical studies (e.g., Culbertson et al., 2010; Karatepe, 2011; Kim and Lee, 2013; Leung et al., 2011) in order to control the risk of CMB. Third, we constructed a psychological separation in the sur- vey by separating PsyCap items from those of work engagement (Time 1) and job satisfaction items from those of affective organi- zational commitment (Time 2) to lower respondents’ perception of any direct connection among the study variables. We also inserted questions about the demographic profile of respondents between PsyCap and work engagement measure sections (Time 1). These five sections of variable items and the questions about demo- graphic profile of respondents, appearing on different pages of the questionnaire, may yield a psychological separation effect on the respondents (Podsakoff et al., 2003). Lastly, after collecting the data, Harman’s 1-factor (often referred to as single-factor) test via an exploratory factor analysis (EFA) and a confirmatory factor analysis (CFA) as a post hoc cor- rective test were used to check the possibility of CMB through the approach excellently outlined by Podsakoff and Organ (1986) and Podsakoff et al. (2003). This approach is one of the most widely used for checking CMB in the hospitality literature (e.g., Kim and Lee, 2013; Mattila and Enz, 2002; Yang and Lau, 2015) as well as other literatures (e.g., Aulakh and Gencturk, 2000; Culbertson et al., 2010; Mossholder et al., 1998).

3.2. Measures

3.2.1. Time 1 survey

PsyCap and work engagement were assessed at Time 1. PsyCap was measured using the 24-item instrument developed by Luthans et al. (2007a). This measures PsyCap as a second-order factor com- prised of the four first-order factors: self-efficacy, optimism, hope, and resilience. This instrument has been validated across multi- ple samples and consistently demonstrates a strong psychometric fit to the data when modeled as a second-order factor in which

each item is fitted to its latent construct and each of the four latent constructs is fitted to an overall PsyCap construct (e.g., Avey et al., 2009; Luthans et al., 2007b). The instrument comprises six items for measuring each of the four factors, with responses collected using

a 7-point Likert-type scale ranging from 1 (“strongly disagree”) to

7 (“strongly agree”). Sample items include “I feel confident analyz- ing a long-term problem to find a solution” (self-efficacy); “When things are uncertain for me at work I usually expect the best” (opti- mism); “If I find myself in a jam at work, I can think of many ways to get out of it” (hope); and “When I have a setback at work, I have trouble recovering from it and moving on” (resilience). The short form of the Utrecht Work Engagement Scale (UWES- 9) developed and validated by Schaufeli et al. (2006) was used to assess work engagement. The UWES-9 consists of nine items rated on a 7-point frequency rating scale ranging from 0 (“never”) to 6 (“always”). The UWES-9 that tapping vigor, dedication, and absorp- tion has sound psychometric properties and an examination of the current literature in hospitality as well as other sectors also pro- vides support for the empirical use of the UWES-9 (e.g., Karatepe, 2011, 2014a,b; Karatepe and Demir, 2014; Sulea et al., 2012). Sam- ple items include “At my work, I feel like I am bursting with energy”;

“I am enthusiastic about my job”; and “I feel happy when I am working intensely.”

3.2.2. Time 2 survey

Job satisfaction and affective organizational commitment as employee morale were measured at Time 2. Job satisfaction was operationalized using eight items from Hartline and Ferrell, 1996, with responses collected using a 7-point Likert-type scale ranging from 1 (“extremely dissatisfied”) to 7 (“extremely satisfied”). A sample item is “The opportunities for advancement with this hotel.”

Affective organizational commitment was operationalized using Allen and Meyer’s (1990) 8-item scale, with responses col- lected using a 7-point Likert-type scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). A sample item is “I would be very happy to spend the rest of my career with this organization.”

3.3. Data collection and screening

Following the suggestions of Podsakoff and colleagues (Podsakoff and Organ, 1986; Podsakoff et al., 2003), the data in this study were collected from front-line hotel employees through a two-wave data collection procedure in order to reduce the risk of CMB. In the first-wave survey (Time 1: February 2013), employees self-reported their PsyCap and work engagement and also provided general profile information such as gender, age, education level, department, and job position. In the second-wave survey (Time 2: March 2013), employees who had completed the first-wave questionnaires were surveyed again and asked to rate their job satisfaction and affective organizational commitment. On completion of a survey questionnaire, which took about five minutes at Time 1 and two at Time 2, the respondent received a

small gift. Data were collected from a sample of front-line employees of 15 five-star hotels in Seoul, South Korea. According to informa- tion received from the Korean Hotel Association at the time the fieldwork was carried out, there were 44 such hotels in Seoul. The researchers contacted the human resource managers of all 44 and 15 hotels allowed the team access to survey their employees. Partic- ipation was voluntary and employees were given assurance of their anonymity and the confidentiality of their answers. Data were col- lected via the following procedures. The human resource managers distributed questionnaires to potential respondents in each depart- ment of their hotel. Each survey packet included a cover letter explaining the purpose of the study. The human resource man- agers undertook to ensure the confidentiality of the data collected. Respondents put their completed surveys into sealed envelopes and gave them to the department managers. The questionnaires were then returned to the researchers. Respondents who had par- ticipated in Time 1 were subsequently asked to take part in the Time

2 survey. Paying the utmost attention to the issue of confidential-

ity, a master list containing the name of each employee and his/her department in the hotel was prepared and controlled by corre- sponding managers only. An identification number was assigned to each employee in this list and was written on each questionnaire instead of the employee’s name. The same procedure was used in the second round of data collection, and the questionnaires at Time

1 and Time 2 were matched using the identification numbers. At Time 1, 400 questionnaires were distributed in the participat- ing hotels, of which 361 were returned, giving a response rate of 90.3%. Fifteen incomplete questionnaires were discarded, leaving 346 for use in the analysis, resulting in an effective response rate of 86.5%. At this stage, we prepared the list of the 346 respondents as described above, and coded the survey questionnaires for Time

2 before distribution. When the Time 2 survey was conducted, the

questionnaires were distributed to these 346 employees, and 340 were completed and returned.

16

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Table 2 Profile of the respondents (n = 312).

Characteristics

Frequency (n)

Percentage (%)

Gender

Male

132

42.3

Female

180

57.7

Age (years)

Under 30

172

55.1

30-39

114

36.5

40 or older

26

8.3

Education level

High school

14

4.5

Two-year college

175

56.1

University

101

32.4

Graduate school

22

7.1

Department Room Food and beverage Others

Job position Rank and file level Supervisor level Assistant manager level Manager level or above

139

44.6

160

51.3

13

4.2

134

63.2

53

25

15

7.1

10

4.7

The dataset was examined for missing data and outliers before conducting the data analysis. It is important to identify outliers because the outliers may bias the mean and influence the normal distribution (Field and Hole, 2003). Among the 340 respondents, after the deletion of missing data and outliers via the Mahalanobis distance, which is a statistical measure of the extent to which cases are multivariate outliers (Byrne, 2001), 28 questionnaires were dis- carded, leaving 312 usable surveys for the Time 2 data analysis. Hence, the final sample of this study consists of 312 front-line hotel employees. The effective response rate for the overall sample was 78%. Table 2 shows the demographic profile of the respondents.

3.4. Data analytic strategy

Data collected from the sample were analyzed using SPSS 21.0 and AMOS 18.0 software. SPSS was used to generate the descrip- tive and inferential statistics. AMOS was used to conduct two-stage structural equation modeling (SEM) to test the hypothesized rela- tionships (Anderson and Gerbing, 1988). The measurement model was first estimated via a confirmatory factory analysis (CFA), and a structural model was then analyzed for model evaluation and research hypotheses testing. The SEM technique is considered more rigorous than typical stepwise regression techniques as all medi- ation paths are measured simultaneously rather than step by step (Clapp-Smith et al., 2009; Hair et al., 2010). In the first step, we first used a single CFA model to assess the higher-order (i.e., second-order) factor structure (with the four first-order factors of self-efficacy, optimism, hope, and resilience) of the PsyCap construct. Then, three first-order factors—work engagement, job satisfaction, and affective organiza- tional commitment—were incorporated into the first CFA model to form a full measurement model (Cha et al., 2013; Hair et al., 2010). In the full measurement model validation stage, the remaining measurement items after the second-order factor structure vali- dation of the PsyCap construct were subjected to CFA to assess the measurement models in terms of reliability and construct valid- ity including convergent, discriminant, and nomological validity (Anderson and Gerbing, 1988; Bagozzi and Yi, 1988; Fornell and Larcker, 1981; Hair et al., 2010). Convergent validity is agreement between measures of the same construct assessed by different methods (Churchill, 1979). To assess convergent validity, the three

criteria of standard factor loading, composite reliability, and aver- age variance extracted (AVE) value were applied (Anderson and Gerbing, 1988; Fornell and Larcker, 1981; Hair et al., 2010). Dis- criminant validity is another important aspect of construct validity that examines the constructs of a model to see if they are distinct from one another (Campbell and Fiske, 1959; Churchill, 1979). This study applied the four criteria of pairwise correlation, correlation confidence interval, 2 difference, and AVE test to assess discrim- inant validity (Anderson and Gerbing, 1988; Fornell and Larcker, 1981; Hair et al., 2010; Kline, 2010). The final common criterion for construct validity is nomological validity, which refers to the extent to which the scale correlates in theoretically predicted ways with measures of different but related constructs (Bagozzi, 1980). In this study, the correlations among variables were estimated to establish the nomological validity (Hair et al., 2010; Netemeyer et al., 1991). CMB is a potentially serious threat, especially in a single-informant survey like this (Podsakoff et al., 2003). Therefore, during the full measurement model analysis, we statistically tested the data to check the potential influence of CMB using Harman’s 1-factor test via EFA and CFA (Podsakoff and Organ, 1986; Podsakoff et al., 2003). In the second step, the hypothesized relationships in the struc- tural model were tested using SEM. We used the four facets (i.e., four first-order factors: self-efficacy, optimism, hope, and resilience) as indicators of the second-order latent construct labeled PsyCap (e.g., Culbertson et al., 2010; Karatepe and Karadas, 2014). In this study, domain-representative parceling (Bandalos and Finney, 2001; Hall et al., 1999;Williams and O’Boyle, 2008) was used to create four parcels by computing the average item scores of self-efficacy, optimism, hope, and resilience as indicators of the PsyCap construct. The use of item parceling (i.e., summated items produced by averaging item scores) helps considerably reduce the number of free parameters, which makes the estimation reliable without increasing the sample size (Bagozzi and Edwards, 1998). To investigate the mediating role of work engagement, both the Preacher and Hayes (2004) bias-corrected bootstrapping method and the Aroian version of the Sobel test (Z-test) 1 suggested by Baron and Kenny (1986) were used. The bootstrapping method is

a more appropriate and more powerful test for the significance of

a mediating or indirect effect than Baron and Kenny’s (1986) tradi-

tional causal steps and product-of-coefficients such as the Sobel test (Sobel, 1982) because it uses a resampling procedure (i.e., the bootstrapping method is a nonparametric approach based on resampling with replacement) to create a confidence interval (CI) for a mediating or indirect effect (MacKinnon et al., 2002; Preacher and Hayes, 2004). This approach is therefore a more popular sta- tistical method for estimating the extent of a mediating or indirect effect than the above-mentioned traditional methods (Bollen and Stine, 1990; Shrout and Bolger, 2002). The combined approach of applying the bootstrapping method and the Aroian version of the Sobel test to mediation analysis is more appropriate than the tra- ditional single method (MacKinnon et al., 2002). In the measurement (both the second-order factor structure of PsyCap and the full measurement models) and structural model estimation stages, the current study used fit indices of the overall 2

measure, normed 2 , goodness of fit index (GFI), root mean square

1 To investigate the mediating effect of work engagement in the relationship between PsyCap and employee morale (job satisfaction and affective organizational commitment), the Aroian version of the Sobel test (Baron and Kenny, 1986) was

a 2 × SE h 2 , where a is the unstan-

dardized path coefficient for the relationship between independent (i.e., PsyCap) and mediator variables (i.e., work engagement); b is the unstandardized path coefficient for the relationship between mediator and dependent variables (i.e., job satisfaction and affective organizational commitment); SE a is the standard error of the relation- ship between independent and mediator variables; and SE b is the standard error of the relationship between mediator and dependent variables.

used: Z value = a × b/ b 2 × SE

2 a + a 2 × SE h 2 + SE

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

17

error of approximation (RMSEA) with 90% CI (HI and LO), normed fit index (NFI), comparative fit index (CFI), incremental fit index (IFI), and Tucker-Lewis index (TLI) to assess model fit (Bollen, 1989; Byrne, 2001; Hair et al., 2010). In addition, the Akaike information criterion (AIC) was used to assess model parsimony to compare competing models (Akaike, 1987; Rust et al., 1995) in both PsyCap factor structure and full measurement models.

4. Results

4.1. Measurement model estimation

4.1.1. Second-order factor structure validation: PsyCap

Prior to estimating the structural model, a CFA was conducted to estimate the measurement model with the maximum likelihood estimation method. Two steps were taken in the validation process of the study variables. First, to confirm the expected second-order factor structure of the PsyCap construct, we began the CFA by fitting this model with six items for each facet (i.e., self-efficacy, opti- mism, hope, and resilience) and then fit each of the four facets to the second-order PsyCap (Table 3). A normed 2 lower than 3.0 (Hair et al., 2010), GFI, NFI, CFI, IFI, and TLI values higher than .90 (Byrne, 2001; Hair et al., 2010), and RMSEA values lower than .08 (Bollen, 1989) all indicate an acceptable model fit. The results of the preliminary CFA thus indicated that the second- order factor structure of PsyCap did not fit the data well with 2 = 934.949, df = 249; Normed 2 = 3.755, GFI = .814, RMSEA [90% CI: LO; HI] = .094 [.088; .101], NFI = .842, CFI = .878, IFI = .879, and TLI = .865. Inspection of the modification indices and correlation measurement errors resulted in respecification of the second-order factor structure of PsyCap. A careful examination of the results of the second-order factor structure of PsyCap therefore suggested deletion of one item each from the self-efficacy, optimism, hope, and resilience measures due to correlation measurement errors. As shown in Table 3, the results of the revised measurement model of the second-order factor structure of PsyCap consisting of 20 items (out of 24), showed improved and good fit with the data ( 2 = 415.677, df = 166; Normed 2 = 2.504; GFI = .883; RMSEA [90% CI: LO; HI] = .070 [.061; .078]; NFI = .908; CFI = .942; IFI = .943; and TLI = .934), as the values of the fit indices were above the model adaptability standard suggested in the literature (Bollen, 1989; Byrne, 2001; Hair et al., 2010). As Table 3 shows, in addition to the hypothesized second-order factor structure for the overall PsyCap measure, we conducted competing model analysis to examine more directly the proposi- tion that PsyCap may be an underlying construct described as a second-order factor structure. The hypothesized second-order fac- tor model described above and the competing 3-, 2-, and 1-factor models were subjected to a significance test of difference using 2 . Specifically, we compared the hypothesized second-order fac- tor model with each of the four facets (i.e., self-efficacy, optimism, hope, and resilience) loading to the second-order factor against 11 competing models including multiple 3- and 2-factor models, which combined various facets as well as a single-factor model in which all items were loaded to one latent PsyCap factor. As shown in Table 3, the hypothesized second-order factor model fits the data better than the 3-, 2-, and 1-factor competing models. For example, the second-order 4-factor model of PsyCap ( 2 = 415.677, df = 166; Normed 2 = 2.504; GFI = .883; RMSEA [90% CI: LO; HI] = .070 [.061; .078]; NFI = .908; CFI = .942; IFI = .943; TLI = .934; and AIC = 503.677) showed further significant improvement in fit ( 2 [1] = 269.744, p < .001), compared to the 3-factor model (i.e., the best fitting of the competing models; Model 5: three factors as indicators of PsyCap:

optimism and hope merged, self-efficacy, and resilience) of Psy- Cap ( 2 = 685.421, df = 167; Normed 2 = 4.104; GFI = .787; RMSEA

[90% CI: LO; HI] = .100 [.092; .108]; NFI = .849; CFI = .880; IFI = .881; TLI = .864; and AIC = 771.421). None of the competing models fit the data well. The results of these model comparisons therefore strongly supported the proposed second-order factor structure of PsyCap as supported in previous studies (e.g., Avey et al., 2009; Luthans et al., 2007b).

4.1.2. Full measurement model validation

After confirming the second-order factor structure of PsyCap, the full measurement model consisting of seven factors and 45 items (self-efficacy: 5; optimism: 5; hope: 5; resilience: 5; work engagement: 9; job satisfaction: 8; and affective organizational commitment: 8) was tested to identify issues of dimensionality, convergent, discriminant, and nomological validity (Anderson and Gerbing, 1988; Bagozzi and Yi, 1988; Fornell and Larcker, 1981; Hair et al., 2010). As can be seen in Table 4, the fit results of the full initial measurement model suggested that the indices fit the data almost as poorly ( 2 = 2220.987, df = 882; Normed 2 = 2.518; GFI = .737; RMSEA [90% CI: LO; HI] = .070 [.066; .073]; NFI = .800; CFI = .868; IFI = .869; and TLI = .859). However, inspection of the modification indices and correlation measurement errors resulted in respecification of the full measurement model. Therefore, a care- ful examination of the results of CFA suggested deletion of one item from work engagement and three items each from job satisfaction and affective organizational commitment measures, due to corre- lation measurement errors. After seven items were deleted from the full initial measurement model, the full revised measurement model, which includes seven factors and 38 items (self-efficacy: 5; optimism; 5 hope: 5; resilience: 5; work engagement: 8; job sat- isfaction: 5; and affective organizational commitment: 5), showed improved and acceptable fit with the data ( 2 = 1375.475, df = 644; Normed 2 = 2.136; GFI = .809; RMSEA [90% CI: LO; HI] = .060 [.056; .066]; NFI = .849; CFI = .913; IFI = .914; and TLI = .905), as the values of the fit indices (with the exception of GFI and NFI) were within the model adaptability standard suggested in the literature (Bollen, 1989; Byrne, 2001; Hair et al., 2010). As shown in Table 4, all the coefficient alphas of the variables exceeded .70, the threshold typically proposed in the literature (self-efficacy: .903; optimism: .885; hope: .891; resilience: .909; work engagement: .909; job satisfaction: .874; and affective orga- nizational commitment: .875) (Hair et al., 2010; Nunnally, 1978), which means that all of the items reflected the variable well (Kline, 2010). All items were significantly linked to their corresponding latent factor (p < .001), with standardized factor loading from .579 to .867, which means that all items effectively measured their corresponding construct (Anderson and Gerbing, 1988). All the composite reliability estimates of the variables exceeded the mini- mum threshold of .70 (self-efficacy: .898; optimism: .864; hope:

.865; resilience: .905; work engagement: .909; job satisfaction:

.866; and affective organizational commitment: .860) (Bagozzi and Yi, 1988; Fornell and Larcker, 1981), indicating that the items of each variables were internally consistent and reliable. The AVE val- ues of self-efficacy, optimism, hope, resilience, work engagement, job satisfaction; and affective organizational commitment were .637, .562, .561, .655, .556, .568, and .554, respectively. These values exceeded the cutoff value of .50 and indicated that at least 55.4% of the variance observed in the items could be accounted for by their hypothesized variables (Fornell and Larcker, 1981; Hair et al., 2010). These findings indicate that the items measuring all vari- ables in the full measurement model were one-dimensional and that within-method convergent validity had been achieved satis- factorily (Anderson and Gerbing 1988; Bagozzi and Yi, 1988; Fornell and Larcker 1981; Hair et al., 2010). As can be seen in Table 5, we performed a CFA of the 6- , 5-, 4-, 3-, 2-, and 1-factor measurement models to compare with the hypothesized 7-factor measurement model. Based on

18

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Table 3 Result of CFA model comparisons: PsyCap factor structure model.

Models

Factors

2

df

Normed

2

df

GFI

RMSEA

NFI

CFI

IFI

TLI

AIC

Model

 

2

[90% CI:

comparison

 

LO; HI]

Model 1

4 factors as indicators of PsyCap Self-efficacy

415.677

166

2.504

0.883

0.07

0.908

0.942

0.943

0.934

503.677

(Hypothesized)

 
 

[.061;

Optimism Hope Resilience 3 factors as indicators of PsyCap Self-efficacy and optimism merged Hope Resilience 3 factors as indicators of PsyCap Self-efficacy and hope merged Optimism Resilience 3 factors as indicators of PsyCap Self-efficacy and resilience merged Optimism Hope 3 factors as indicators of PsyCap Optimism and hope merged Self-efficacy Resilience 3 factors as indicators of PsyCap Optimism and resilience merged Self-efficacy Hope 3 factors as indicators of PsyCap Hope and resilience merged Self-efficacy Optimism 2 factors as indicators of PsyCap Optimism, hope, and resilience merged Self-efficacy 2 factors as indicators of PsyCap Self-efficacy, hope, and resilience merged Optimism 2 factors as indicators of PsyCap Self-efficacy, optimism, and resilience merged Hope 2 factors as indicators of PsyCap Self-efficacy, optimism, and hope merged Resilience 1 factor as an indicator of PsyCap All 20 indicators

.078]

Model 2

898.436

167

5.38

482.759

1

0.698

0.119

0.801

0.831

0.832

0.808

984.436

1 and 2

 

[.111;

.126]

Model 3

797.091

167

4.773

381.414

1

0.735

0.11

824

0.855

0.855

0.835

883.091

1 and 3

 

[.103;

.118]

Model 4

998.43

167

5.973

582.753

1

0.664

0.127

0.779

0.808

0.809

0.782

1084.43

1 and 4

 

[.119;

.134]

Model 5

685.421

167

4.104

269.744

1

0.787

0.1

0.849

0.88

0.881

0.864

771.421

1 and 5

 

[.092;

.108]

Model 6

686.77

167

4.112

271.093

1

0.779

0.1

0.848

0.88

0.881

0.864

772.77

1 and 6

 

[.092;

.108]

Model 7

825.199

167

4.941

409.522

1

0.724

0.113

0.818

0.848

0.849

0.827

911.199

1 and 7

 

[.105;

.120]

Model 8

1006.743

169

5.957

591.066

3

0.695

0.126

0.778

0.807

0.808

0.783

1088.743

1 and 8

 

[.119;

.134]

Model 9

1271.344

169

7.523

855.667

3

0.62

0.145

0.719

0.746

0.747

0.714

1353.344

1 and 9

 

[.137;

.152]

Model 10

1222.478

169

7.234

806.801

3

0.635

0.142

0.73

0.757

0.758

0.727

1304.478

1 and

 

10

 

[.134;

.149]

Model 11

1116.479

169

6.606

700.793

3

0.666

0.134

0.753

0.781

0.783

0.754

1198.479

1 and

 

11

 

[.127;

.142]

Model 12

1479.307

170

8.702

1063.63

4

0.592

0.157

0.673

0.698

0.699

0.663

1559.307

1 and

 

12

 

[.150,

 

.165]

Notes: One item each from the measures of self-efficacy (“I feel confident presenting information to a group of colleagues”), optimism (“When things are uncertain for me at work I usually expect the best”), hope (“Right now I see myself as being pretty successful at work”), and resilience (“I usually manage difficulties one way or another at work”) was deleted due to correlation measurement errors in all models. All 2 difference scores among PsyCap factor structure models were significant at p < .001. PsyCap = psychological capital; GFI = goodness of fit index; RMSEA = root mean square error of approximation; NFI = normed fit index; CFI = comparative fit index; IFI = incremental fit index; TLI = Tucker-Lewis index; AIC = Akaike information criterion; and CI = confidence interval.

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

19

Table 4 Results of reliability and convergent validity tests: full measurement model.

Variable and items

Standardized

t-Value

Alpha

CR

AVE

loading

PsyCap: self-efficacy SE1. I feel confident analyzing a long-term problem to find a solution SE2. I feel confident in presenting my work area in meetings with management SE3. I feel confident contributing to discussions about my hotel’s strategy SE4. I feel confident helping to set targets/goals in my work area SE5. I feel confident contacting people outside my hotel (e.g., customers) to discuss problems

 

0.903

0.898

0.637

0.863

18.681

0.841

0.778

15.973

0.781

16.081

0.78

16.058

PsyCap: optimism OP2. If something can go wrong for me work-wise, it will R OP3. I always look on the bright side of things regarding my job OP4. I’m optimistic about what will happen to me in the future as it pertains to work OP5. In my job, things never work out the way I want them to R OP6. I approach my job as if every cloud has a silver lining

 

0.885

0.864

0.562

0.723

14.01

0.69

13.179

0.847

17.382

0.831

16.927

0.819

PsyCap: hope HO1. If I find myself in a jam at work, I can think of many ways to get out of it HO2. At the present time, I am energetically pursuing my goals HO3. There are lots of ways around any problem that I am facing now HO5. I can think of many ways to reach my current goals HO6. At this time, I am meeting the work goals I have set for myself

 

0.891

0.865

0.561

0.797

16.102

0.828

0.739

14.503

0.788

15.854

0.786

15.781

PsyCap: resilience RE1. When I have a setback at work, I have trouble recovering from it and moving on R RE3. I can be “on my own,” so to speak, at work if I have to RE4. I usually take stressful things at work in my stride RE5. I can get through difficult times at work because I’ve experienced difficulties before RE6. I feel I can handle many things at a time at my job

 

0.909

0.905

0.655

0.751

15.883

0.849

19.476

0.828

18.631

0.867

0.808

17.895

Work engagement WE1. At my work, I feel like I am bursting with energy WE3. When I get up in the morning, I feel like going to work WE4. I am enthusiastic about my job WE5. My job inspires me WE6. I am proud of the work I do WE7. I feel happy when I am working intensely WE8. I am immersed in my work WE9. I get carried away when I am working

 

0.909

0.909

0.556

0.797

15.045

0.765

14.31

0.753

14.042

0.765

14.309

0.74

13.761

0.773

0.698

12.844

0.707

13.037

Job satisfaction JS1. My overall job JS2. My fellow workers JS3. My supervisor JS4. This hotel’s policies JS5. The support provided by this hotel

 

0.874

0.866

0.568

0.789

15.581

0.831

16.722

0.814

16.236

0.817

0.579

10.566

Affective organizational commitment AOC1. I would be very happy to spend the rest of my career with this hotel AOC3. I really feel as if this hotel’s problems are my own AOC4. I do not feel like part of the family at my hotel R AOC5. I do not feel emotionally attached to this hotel R AOC6. This hotel has a great deal of personal meaning for me

 

0.875

0.86

0.554

0.624

11.775

0.769

15.551

0.811

16.812

0.792

16.231

0.84

Model fit Initial model (45 items): 2 = 2220.987, df = 882; Normed 2 = 2.518; GFI = .737; RMSEA [90% CI: LO; HI] = .070 [.066; .073]; NFI = .800; CFI = .868; IFI = .869; and TLI = .859. Revised model (38 items): 2 = 1375.475, df = 644; Normed 2 = 2.136; GFI = .809; RMSEA [90% CI: LO; HI] = .060 [.056; .066]; NFI = .849; CFI = .913; IFI = .914; and TLI = .905. Notes: PsyCap was manifested by self-efficacy, optimism, hope, and resilience. One item each from self-efficacy (SE6: “I feel confident presenting information to a group of colleagues”), optimism (OP1: “When things are uncertain for me at work I usually expect the best”), hope (HO4: “Right now I see myself as being pretty successful at work”), and resilience (RE2: “I usually manage difficulties one way or another at work”) was discarded during the second-order factor structure validation. One item from the work engagement measure (WE2: “At my job, I feel strong and vigorous”) and three items each form job satisfaction (JS6: “My salary”; JS7: “The opportunities for advancement with this hotel”; and JS8: “This hotel’s customers”), and affective organizational commitment measures (AOC2: “I enjoy discussing my hotel with people outside it”; AOC7:

“I do not feel a strong sense of belonging to my hotel” R; and AOC8. “I think that I could easily become as attached to another hotel as I am to this one” R) were discarded during the convergent validity test. All standardized factor loadings are significant at p < .001. R denotes reverse-scored items. PsyCap = psychological capital; CR = composite reliability; AVE = average variance extracted; GFI = goodness of fit index; RMSEA = root mean square error of approximation; NFI = normed fit index; CFI = comparative fit index; IFI = incremental fit index; and TLI = Tucker–Lewis index.

the fit indices, as hypothesized, the 7-factor measurement model fit the data better than the six competing measurement mod- els (Akaike, 1987; Hair et al., 2010; Rust et al., 1995), with fit indices demonstrating good discriminant validity of the study vari- ables. For instance, the hypothesized 7-factor measurement model

( 2 = 1375.475, df = 644; Normed 2 = 2.136; GFI = .809; RMSEA

[90% CI: LO; HI] = .060 [.056; .066]; NFI = .849; CFI = .913; IFI = .914;

TLI = .905, and AIC = 1569.475) showed that further significant improvement in fit ( 2 [6] = 284.086, p < .001), compared to the 6-factor measurement model (i.e., the best fit model among the

competing models); Model 2: job satisfaction and affective orga- nizational commitment merged; self-efficacy; optimism; hope; resilience; and work engagement ( 2 = 1659.561, df = 650; Normed 2 = 2.553; GFI = .764; RMSEA [90% CI: LO; HI] = .071 [.066; .075]; NFI = .818; CFI = .880; IFI = .881; TLI = .870; and AIC = 1841.561). All competing full measurement models had poor fit. The results from these fullmeasurementmodel comparisons therefore strongly sup- ported the hypothesized 7-factor full measurement model. In the current study, discriminant validity was assessed by four criteria (Anderson and Gerbing, 1988; Fornell and Larcker 1981;

20

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Table 5 Result of CFA model comparisons: full measurement model.

Models

Factors

2

df

Normed

2

df

GFI

RMSEA

NFI

CFI

IFI

TLI

AIC

Model

 

2

[90% CI:

comparison

 

LO; HI]

Model 1

7 variables served as factors Self-efficacy

1375.475

644

2.136

0.809

0.06

0.849

0.913

0.914

0.905

1569.475

(Hypothesized)

 
 

[.056;

 

.066]

 

Optimism

Hope

Resilience

Work engagement

Job satisfaction

Affective

organizational

commitment

Model 2

6 variables served as factors Job satisfaction and affective organizational commitment merged Self-efficacy Optimism Hope Resilience Work engagement

1659.561

650

2.553

284.086

6

0.764

0.071

0.818

0.88

0.881

0.87

1841.561

1 and 2

 

[.066;

.075]

Model 3

5 variables served as factors Work engagement, job satisfaction, and affective organizational commitment merged Self-efficacy Optimism Hope Resilience

1826.561

655

2.789

451.086

11

0.749

0.076

0.8

0.861

0.862

0.851

1998.561

1 and 3

 

[.072;

.080]

Model 4

4 variables served as factors Resilience, work engagement, job satisfaction, and affective organizational commitment merged Self-efficacy Optimism Hope

2222.067

659

3.372

846.592

15

0.698

0.087

0.756

0.814

0.815

0.802

2386.067

1 and 4

 

[.083;

.091]

Model 5

3 variables served as factors Hope, resilience, work engagement, job satisfaction, and affective organizational commitment merged Self-efficacy Optimism

2532.089

662

3.825

1156.614 18

0.663

0.095

0.723

0.778

0.779

0.764

2690.089

1 and 5

 

[.091;

.099]

Model 6

2 variables served as factors Optimism, hope, resilience, work engagement, job satisfaction, and affective organizational commitment merged Self-efficacy

2788.205

664

4.199

1412.73

20

0.633

0.101

0.694

0.748

0.749

0.733

2942.205

1 and 6

 

[.098;

.105]

Model 7

1 variable served as a factor All seven study variables merged

3298.247

665

4.96

1922.772 21

0.579

0.113

0.639

0.687

0.689

0.669

3450.247

1 and 7

 

[.109;

.117]

Notes: All 2 difference scores among CFA models are significant at p < .001. GFI = goodness of fit index; RMSEA = root mean square error of approximation; NFI = normed fit index; CFI = comparative fit index; IFI = incremental fit index; TLI = Tucker–Lewis index; AIC = Akaike information criterion; and CI = confidence interval.

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

21

Table 6 Mean, standard deviation, and the results of nomological and discriminant validity tests.

Variables

M

SD

1234567

 

1. PsyCap: self-efficacy

5.167

.880

[.637]

(42.861)

(23.396)

(37.701)

(47.500)

(41.350)

(40.632)

 

.294

.347

.261

.278

.237

.275

2. PsyCap: optimism

4.924

.850

.542

[.562]

(20.970)

(19.454)

(32.371)

(38.605)

(27.429)

 

[.658, .426]

.416

.457

.426

.285

.371

3. PsyCap: hope

4.956

.947

.589

.645

[.561]

(19.833)

(22.237)

(20.844)

(16.559)

 

[.725, .453]

[.785, .505]

.340

.399

.349

.379

4. PsyCap: resilience

5.131

.894

.511

.676

.583

[.655]

(23.730)

(25.187)

(21.177)

 

[.635, .387]

[.812, .540]

[.727, .439]

.441

.356

.372

5. Work engagement

4.091

.794

.527

.653

.632

.664

[.556]

(24.443)

(21.583)

 

[.639, .415]

[.775, .531]

[.768, .496]

[.796, .532]

.482

.500

6. Job satisfaction

5.085

.863

.487

.534

.591

.597

.694

[.568]

(25.193)

 

[.605, .369]

[.654, .414]

[.731, .451]

[.731, .463]

[.824, .564]

.399

7. Affective

5.074

.882

.524

.609

.616

.610

.707

.632

[.554]

organizational

[.644, .404]

[.737, .481]

[.762, .470]

[.748, .472]

[.841, .573]

[.764, .500]

commitment

Notes: Composite scores for each measure were obtained by averaging scores across remaining items after reliability and convergent validity tests representing that measure. PsyCap was manifested by self-efficacy, optimism, hope, and resilience. Correlations among all study variables are presented in the lower off diagonal; confidence interval of plus or minus two standard errors around the correlation among study variables are presented in parentheses in the lower off diagonal; calculated values of the shared variance (i.e., squared correlations) among study variables are presented in the upper off diagonal; the 2 difference scores among study variables are presented in parentheses in the upper off diagonal; and AVE values (boldface type) are presented in parentheses along the diagonal. All correlations among study variables are significant at p < . 01. All 2 difference scores among study variables (all tests = 1df): 2 > 10.830, p < .001. PsyCap = psychological capital; and AVE = average variance extracted.

Hair et al., 2010; Kline, 2010): pairwise correlation, correlation con- fidence interval, 2 difference, and AVE tests (Table 6). First, to evaluate the distinctiveness of the variables measured by differ- ent sets of items, pairwise correlations between latent variables were checked. The highest pairwise correlation value was .707 (between work engagement and affective organizational commit- ment), which does not exceed the maximum threshold of .850 suggested by Kline (2010), indicating that the different variables were measured by different sets of items. Second, we calculated the confidence interval of plus or minus two standard errors around the correlation between the variables and determined whether this interval includes 1.0. If it does not include 1.0, discriminant validity is demonstrated (Anderson and Gerbing, 1988). As can be seen in Table 6, none of the 21 con- fidence intervals (e.g., between self-efficacy and optimism: .658; .426) include 1.0 in this study. Third, the measurement model was constrained by forcing a correlation of 1.0 between a pair of two variables, the constrained model was then estimated, and the results were compared to the free model using a 2 difference test. Twenty-one such correlations and estimates were executed. In all cases, the resulting 2 value of the constrained model was greater than that of the free model, and the 2 value was significant at p < .001 (the lowest 2 differ- ence score was achieved between hope and affective organizational commitment; the constrained model: 2 = 123.087, df = 35; the free model: 2 = 106.528, df = 34; 2 (1) = 16.559, p < .001), which indicates evidence of the discriminant validity for study variables (Anderson and Gerbing, 1988). Lastly, we compared the AVEs for the two variables of interest with the square of the correlation between them. This approach is a more rigorous test than the above three approaches (Hair et al., 2010). Discriminant validity is demonstrated if the AVEs are greater than the corresponding squared correlation (Fornell and Larcker 1981; Hair et al., 2010). All squared correlations among the variables were smaller than the AVEs of the respective variables (e.g., between self-efficacy and optimism: r 2 = .294; AVE of self- efficacy = .637, AVE of optimism = .562). As can be seen in Table 6, the results using the four criteria therefore indicate strong evidence of the discriminant validity of the study variables in this study (Anderson and Gerbing, 1988; Fornell and Larcker 1981; Hair et al., 2010; Kline, 2010). The correlation estimates calculated from the validation sam- ple are shown in Table 6. Because PsyCap variable is a second-order

construct, nomological validity would be established if the scores of the measures of the four first-order factors (i.e., self-efficacy, opti- mism, hope, and resilience) of PsyCap positively and significantly correlated with work engagement, job satisfaction, and affective organizational commitment, as hypothesized in Hypotheses 1–3. In addition, nomological validity would be established if the scores of the measures of work engagement positively and significantly cor- related job satisfaction and affective organizational commitment, as hypothesized in Hypotheses 4 and 5. All correlations between the four first-order factors of PsyCap and work engagement (e.g., self-efficacy and work engagement: r = .527)/job satisfaction (e.g., self-efficacy and job satisfaction: r = .487)/affective organiza- tional commitment (e.g., self-efficacy and affective organizational commitment: r = .524), and between work engagement and job satisfaction (r = .694)/affective organizational commitment were significant (r = .707) at p < .01, which indicates strong evidence of the nomological validity of the study variables (Hair et al., 2010; Netemeyer et al., 1991).

4.1.3. Common method bias checking

After the full measurement model validation, we performed an EFA with unrotation solution on all the remaining 38 items to check the possibility of common method problems. The CMB is present when a single factor is yielded from the factor analysis or one gen- eral factor accounts for more than 50% of the covariance among the variables (Mattila and Enz, 2002; Podsakoff and Organ, 1986). The results of factor analysis revealed a 7-factor structure (with eigen- values greater than 1.0), with each factor accounting for less than 50% of the covariance among the variables. In addition, we applied a CFA to compare a model with a single latent factor to our 7-factor model. The basic assumption of this method is that if a substantial amount of CMB is present, a single latent factor will fit the data well (Mossholder et al., 1998; Podsakoff et al., 2003). The 1-factor model, in which all the remaining 38 items after the full measure- ment model validation load on a single factor, has a relatively very poor fit to the data with 2 = 3298.247, df = 665; Normed 2 = 4.960; GFI = .579; RMSEA [90% CI: LO; HI] = .113 [.109; .117]; NFI = .639; CFI = .687; IFI = .689; and TLI = .669. A 2 difference test furthermore indicated that the 7-factor model was superior to the single-factor model, 2 (21) = 1922.772, p < .001. The results allow us to con- clude that the common method problem is unlikely to be pervasive in the current study (Podsakoff and Organ, 1986; Podsakoff et al.,

2003).

22

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

R 2 = .631 Hypothesis 2: (.398, 3.701)*** Job Satisfaction Self-efficacy Hypothesis 6: Partial mediating
R 2 = .631
Hypothesis 2:
(.398, 3.701)***
Job Satisfaction
Self-efficacy
Hypothesis 6: Partial mediating role of work engagement
Indirect effect: PsyCap → work engagement → job satisfaction
Point estimate: .363**; bias-corrected bootstrap 90% CI: .194 (LL); .536 (UL)
The Aroian version of the Sobel test: Z = 3.823***
Time 2
(.678, 12.346)***
R 2 = .715
Hypothesis 4:
Hypothesis 1:
Optimism
(.429, 4.026)***
(.845, 12.474)***
[.459, .114]
[.941, .075]
PsyCap
Work Engagement
(.813, 15.371)***
(.789)***
Time 1
Time 1
Hypothesis 5:
Hope
(.336, 3.306)***
[.347, .105]
(.794, 14.936)***
R 2 = .701
Resilience
Hypothesis 7: Partial mediating role of work engagement
Indirect effect: PsyCap → work engagement → affective organizational commitment
Point estimate: .284*; bias-corrected bootstrap 90% CI: .077 (LL); .489 (UL)
The Aroian version of the Sobel test: Z = 3.186***
Affective Organizational
Commitment
Time 2
Hypothesis 3:

(.534, 5.017)***

Model fit χ 2 = 438.134, df = 203; Normed χ 2 = 2.158; GFI = .890; RMSEA [90% CI: LO; HI] = .061 [.053; .069]; NFI = .905; CFI = .947; IFI = .947; and TLI = .939.

Fig. 3. Results of structural model estimation and hypotheses test for the effect of PsyCap on employee morale with the mediator of work engagement.

4.2. Structural model estimation and hypotheses testing

As the full measurement model was deemed to be valid and acceptable as well as same source problem is unlikely to be a per- vasive problem in the current study, the structural model was estimated and the hypotheses were tested in the second step (Fig. 3). The fit indices from the SEM demonstrated that the fit of the proposed structural model was satisfactory ( 2 = 438.134, df = 203; Normed 2 = 2.158; GFI = .890; RMSEA [90% CI: LO; HI] = .061 [.053; .069]; NFI = .905; CFI = .947; IFI = .947; and TLI = .939).

4.2.1. Main effects and hypotheses testing

As can be seen in Fig. 3, a significant and positive relation- ship between PsyCap and work engagement was found (ˇ = 0.845,

p < .001), supporting Hypothesis 1. PsyCap explained 71.5% of the variance in work engagement. Consistent with Hypotheses 3 and

4, PsyCap (ˇ = 0.398, p < .001) and work engagement (ˇ = 0.429,

p < .001) both had a significant and positive impact on job satis- faction, jointly explaining 63.1% of the variance in this measure. Finally, the results also provided empirical support for Hypotheses 3 and 5, because PsyCap (ˇ = 0.534, p < .001) and work engagement (ˇ = .336, p < .001) both had a significant and positive impact on affective organizational commitment, jointly explaining 70.1% of the variance in this measure.

4.2.2. Mediating effects and hypotheses testing

As shown in Fig. 3, to investigate the mediating role of work engagement, we utilized both the bias-corrected bootstrapping method (Preacher and Hayes, 2004) and the Aroian version of the Sobel test (Baron and Kenny, 1986). When using the boot- strapped CI (lower limit of the CI [LL]; upper limit of the CI [UL])

procedure, mediation is indicated by the exclusion of zero from the CI for the indirect effect. If the bootstrapped CI does not include zero, it can be said with confidence that the mediating

effect differs from zero. In the current study, 90% bias-corrected

CI was estimated using 5000 bootstrapped samples. The indi-

rect effects of PsyCap work engagement job satisfaction (point estimate of indirect effect = .363, p < .01; CI [LL: .194; UL: .536])

and PsyCap work engagement affective organizational com- mitment (point estimate of indirect effect = .284, p < .05; CI [LL:

.077; UL:.489]) were significant and the 90% CI did not include zero. Following the recommendation of Baron and Kenny (1986), we also used the Aroian version of the Sobel test for the effect of Psy- Cap on job satisfaction via work engagement (Z = 3.823, p < .001), which showed that work engagement had a significant mediating effect on this relationship. Furthermore, the Z-score for the effect of PsyCap on affective organizational commitment via work engage- ment was also significant (Z = 3.186, p < .001), indicating that work engagement also had a mediating effect on this linkage. To summarize, both the Aroian version of the Sobel test and the bias-corrected bootstrapping method indicated that work engage- ment significantly mediates the effects of PsyCap on job satisfaction and affective organizational commitment. Specifically, because the direct effects of PsyCap on both job satisfaction (Hypothesis 2) and affective organizational commitment (Hypothesis 3) were positive and significant, these results demonstrated that work engagement partially mediates the relationships between PsyCap and both job satisfaction and affective organizational commitment. Therefore, Hypotheses 6 and 7 were also supported.

5. Discussion and implications

5.1. Discussion

This study has investigated work engagement as a partial medi- ator of the effect of PsyCap on two employee morale variables—job satisfaction and affective organizational commitment—in the con- text of the hospitality industry.We set out to examine the structural role of work engagement in the linkage between PsyCap and these variables of employee morale. As hypothesized, work engage- ment was shown to be a positive consequence of PsyCap and the antecedent of the two variables of employee morale. At the same time, since PsyCap has a direct impact on these variables, work engagement also acted as a partial mediator of these relationships. The role of work engagement in this linkage can be explained with reference to COR theory (Hobfoll, 1989, 2001) and the JD-R model

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

23

(Bakker and Demerouti, 2007, 2008; Bakker and Leiter, 2010). Based on these theories, it can be argued that service employees who demonstrate a high level of PsyCap will tend to show higher work engagement as a consequence. Previously, the role of PsyCap as a strong predictor of the work engagement level of an employee (i.e., one part of the structure of this study) had only been proposed conceptually (Sweetman and Luthans, 2010). This has now been empirically demonstrated. In addition, this study illustrates the effect of PsyCap not only on work engagement but also on the two employee morale vari- ables. PsyCap acted as a predictor of the variables, with a stronger direct effect on affective organizational commitment than job sat- isfaction. However, PsyCap had a stronger indirect effect on job satisfaction through work engagement than on affective organiza- tional commitment. Therefore, the result of combining the direct and indirect effects of PsyCap on the two employee morale vari- ables indicate that PsyCap had a similarly positive effect on both, given the partial mediation role of work engagement. These results imply that affective organizational commitment relies more on work engagement than job satisfaction, which has not been con- sidered before. Therefore, this study offers new and interesting findings about the structural role of work engagement as it relates to PsyCap and the two employee morale constructs. Moreover, the results of this study demonstrate that PsyCap has a very strong effect on work engagement (ˇ = 0.845). However, the effect of work engagement on the two employee morale variables was smaller (ˇ = 0.398 for job satisfaction and ˇ = 0.534 for affec- tive organizational commitment). It might therefore be presumed that work engagement is more closely related to PsyCap than the employee morale variables. Sweetman and Luthans (2010) theorize that the positive emotions created by the link between PsyCap and work engagement are the foundation of their strong relationship. This result is especially meaningful from the perspective of organi- zational psychology, as well as in the service industry context, with Sweetman and Luthans (2010) having identified PsyCap and work engagement as psychological factors. However, employee morale is rarely considered as a psychological factor. Although previous work suggests that work engagement is promoted by various types of resources which increase job outcome variables (e.g., Karatepe et al., 2013; Kim et al., 2009), somewhat surprisingly, such work has not considered the fact that work engagement, as one of the psy- chological components in a work setting, would be more strongly affected by other psychological factors, such as PsyCap, as shown here.

5.2. Strengths and theoretical contributions

By taking these two critical emerging constructs, PsyCap and work engagement, and linking them in a new way to employee morale as denoted by job satisfaction and affective organizational commitment, this study makes a number of significant theoretical contributions. First, the structural relationships between PsyCap, work engagement, and employee morale as reported here not only confirm the role of work engagement in PsyCap and employee morale (which has not been shown before) but also identify the pivotal role of work engagement as a mediator between PsyCap and employee morale. This is significant, because although PsyCap or work engagement have been highlighted by a wide-ranging body of organizational psychology research, critical questions remain about how these important constructs should be linked with each other and with other variables. Most research has tended to focus on the effect of either PsyCap or work engagement on other con- structs. Such studies have contributed to our understanding of PsyCap (e.g., Avey et al., 2009; Luthans and Youssef, 2007; Nelson and Cooper, 2007) or work engagement (e.g., Bakker and Leiter, 2010), and on defining the antecedents and/or consequences of

one of the two key constructs (e.g., Xanthopoulou et al., 2009; Yalabik et al., 2013). However, only a few studies have explored or conceptually proposed a relationship between PsyCap and work engagement (e.g., Sweetman and Luthans, 2010). This study there- fore breaks new ground by proposing a structural model and giving meaningful consideration to a holistic view of the linkages between PsyCap, work engagement, and other constructs, with a particular focus on the aspects of organizational psychology and behavior/attitude. Second, this study proposes that work engagement is a par- tial mediator of the relationship between PsyCap and the two employee morale variables. Since work engagement has a vari- ety of antecedents and consequent variables, it acts as a critical mediator, and a number of studies have found that it has a full effect (e.g., Karatepe et al., 2013; Karatepe, 2013a,b). However, by proposing work engagement as a partial mediator, this study has mapped out a research framework which suggests there are recip- rocal and simultaneous relationships between a predictor (PsyCap), work engagement, and consequent variables (the two employee morale variables of job satisfaction and affective organizational commitment). This is critical from an organizational psychology perspective, because the partial mediating role of work engage- ment shows the importance of a direct relationship between a predictor of work engagement and the outcome variables, rather than presenting it in a fully mediating role. This implies that the relationships between the predictor, mediator, and outcome con- structs in organizational psychology research could include more complicated reciprocal or simultaneous mechanisms which cannot be explained superficially. Since only a few studies have argued for this conceptually, with most proposing only simple structures, the empirical testing of the structural model presented here offers a meaningful research direction for future work. As mentioned in the discussion above, the effects of PsyCap on the employee morale variables were much stronger than those of work engagement. This provides a very important insight into the role of PsyCap and work engagement as essential antecedents of employee morale. Although work engagement has been high- lighted before as a vital component of boosting employee morale (e.g., Yeh, 2013), PsyCap has a more fundamental function as a cause and antecedent of employee morale. Although both PsyCap and work engagement are critical, as argued previously, our results suggest that PsyCap has a stronger effect on employee morale, implying that the individual capital (or personal resources) of an employee has a more significant role. Many studies have already highlighted the importance of job resources, but these findings emphasize the fundamental significance of personal rather than generic work-related resources. Moreover, they also provide an alter- native explanation for the strong effectiveness of the clustering between PsyCap and work engagement, which might synergisti- cally boost employee morale to a higher level. Finally, this study has not only confirmed the critical roles of Psy- Cap and work engagement in the service industry, but has added employeemorale as an important purposive construct. Generic out- come variables related to job performance, which have been and remain central to the purposes of many companies, have been stud- ied extensively. However, by using the employee morale variables, this study emphasizes a positive view of organizational behav- iors/attitudes, which has been a recent focus of research and better reflects the needs of modern companies.

5.3. Practical implications

The findings of this study also have practical implications for hotel managers seeking to plan and introduce effective tools to help employees develop positive organizational attitudes and behav- iors by deploying their psychological resources; in other words, to

24

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

improve hotel employees’ job satisfaction and affective organiza-

tional commitment. First, it is imperative that service organizations should recognize the prime importance of PsyCap and develop training programs to help service workers develop and maintain it

at high levels (i.e., by sustaining their self-efficacy, hope, optimism,

and resilience), including ways of protecting themselves from the loss of such capital. Staff can protect, increase, or maintain their PsyCap more effectively by learning how to enhance their self-

efficacy, hope, optimism, and resilience. As discussed by Avey et al. (2009), short training interventions can enhance all these dimen- sions, as well as PsyCap overall. Moreover, more comprehensive development programs ranging from PsyCap and work engage- ment to employee morale would be even more effective. Through such programs, service workers can learn how to forecast obstacles; prevent the loss and maintain good levels of PsyCap; experience

a high level of engagement with their job and workplace; feel a

sense of job satisfaction; and enjoy and feel emotionally attached and more committed to their organization. Such experience and recognition will lead them to set higher performance goals. More- over, such development programs can be considered at a group or organization level, since social capital such as supervisor sup- port adds value and increases positive organizational behaviors (Larson and Luthans, 2006). In addition, according to these findings, empowerment and rewards could be made available to enhance the link between work engagement and employee morale. Empower- ment and rewards are recognized as the most crucial indicators of high-performance work practices and are strong influencers to pro- mote work engagement (Karatepe, 2013a). As such, overall human resource management practices can help to create a better working environment, which leads to improved job performance. Second, management should consider the beneficial impact of PsyCap on employees’ positive organizational behaviors and attitudes in the recruitment process. As shown here, an individ- ual’s personal resources, such as psychological factors, are very important to their level of satisfaction and performance. Service organizations could gain an advantage by using a more thorough selection process which aims to hire employees with high lev- els of PsyCap (through means such as developing and using an instrument which assess applicants’ PsyCap levels). Moreover, an emphasis on PsyCap during the recruiting process would deliver the very important message to potential employees that high Psy- Cap is a critical characteristic for service workers, encouraging them to make a continuous effort to maintain and increase their PsyCap once they start work.

5.4. Limitations and future studies

Although this empirical study makes several important con-

tributions to the existing knowledge base, it does have certain limitations which suggest viable prospects for future research. First,

a one-month time-lagged research design was used to collect the

data. Although such an approach can provide some evidence for

temporal causality, it is not enough to make firm causal inferences.

In order to be able to do so, future studies should collect data over a

longer period than was possible here (cf., Grandey and Cropanzano,

1999).

Second, replication studies using larger samples in different hos- pitality settings in Korea and other countries are needed to obtain findings that can be generalized more widely and cross-validated. Third, the results of this study suggest that work engagement acts as a predictor of employee morale as represented by job satisfaction and affective organizational commitment. Jobs in the hospitality sector require employees with the creativity to generate new ideas about work processes, methods, services, and products

(Hon, 2011). Individual innovative behaviors are thus important for maintaining the competitive advantage and long-term success of

hospitality firms. Therefore, an investigation of PsyCap as a booster of service innovation or creativity (cf., Jafri, 2012; Rego et al., 2012) in hospitality employees may yield important insights for both aca- demics and practitioners. Fourth, this study has considered two variables of employee morale as consequences of work engagement and job outcomes. Although the findings have confirmed the role of work engagement as a precedent of the employee morale variables, other studies view job satisfaction and affective organizational commitment as vari- ables denoting employee attitudes and suggest instead that they are outcome variables of work engagement (Yalabik et al., 2013). Other work proposes a mutual relationship among the resources dimen- sions and work engagement (Xanthopoulou et al., 2009), implying that the relationships among these constructs are complicated. Future empirical studies could investigate such a sub-mechanism, not incorporated in the structural model of this study, in order to examine other structural relationships among these constructs. Fifth, this study has focused on the role of PsyCap as a predic- tor of employee morale, without using a moderator. However, a few studies have examined the moderating role of PsyCap on the relationship between leader-member exchange and follower job performance (e.g., Wang et al., 2012) and between perceptions of organizational politics, turnover intention, job satisfaction, and job performance (e.g., Abbas et al., 2014). Service recovery is a cru- cial aspect of job-related performance in the hospitality context. On the basis of this line of research, it would be useful to incor- porate PsyCap into the extended model proposed as a form of personal resource, in order to understand whether it can moderate the impact of employee morale on service recovery performance. Finally, another avenue for consideration may be business- unit level relationships rather than those between individuals. Harter et al. (2002) investigate the business-unit level relationship between employee satisfaction, employee engagement, and busi- ness outcomes. Collective measurement at this level would provide interesting insights into how to manage the constructs considered in this study more comprehensively across an organization.

References

Abbas, M., Raja, U., Darr, W., Bouckenooghe, D., 2014. Combined effects of perceived politics and psychological capital on job satisfaction, turnover intentions, and performance. J. Manage. 40 (7), 1813–1830. Akaike, H., 1987. Factor analysis and AIC. Psychometrika 52 (3), 317–332. Allen, N.J., Meyer, J.P., 1990. The measurement and antecedents of affective, continuance, and normative commitment to the organization. J. Occup. Psychol. 63 (1), 1–8. Ambrose, M.L., Arnaud, A., Schminke, M., 2008. Individual moral development and ethical climate: the influence of person-organization fit on job attitudes. J. Bus. Ethics 77 (3), 323–333. Anderson, J.C., Gerbing, D.W., 1988. Structural equation modeling in practice: a review and recommended two-step approach. Psychol. Bull. 103 (3), 411–423. Anheier, H.K., Gerhards, J., Romo, F.P., 1995. Forms of capital and social structure in cultural fields: examining Bourdieu’s social topography. Am. J. Sociol. 100 (4),

859–903.

Aulakh, P.S., Gencturk, E.F., 2000. International principal-agent relationships:

control, governance and performance. Ind. Mark. Manage. 29 (6), 521–538. Avey, J.B., Luthans, F., Jensen, S.M., 2009. Psychological capital: a positive resource for combating employee stress and turnover. Hum. Res. Manage. 48 (5),

677–693.

Bagozzi, R.P., 1980. Causal Models in Marketing. John Wiley & Sons, New York, NY. Bagozzi, R.P., Edwards, J.R., 1998. A general approach for representing constructs in organizational research. Organ. Res. Methods 1 (1), 45–87. Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. J. Acad. Mark. Sci. 16 (1), 74–94. Bakker, A.B., Demerouti, E., 2007. The job demands-resources model: state of the art. J. Manage. Psychol. 22 (3), 309–328. Bakker, A.B., Demerouti, E., 2008. Towards a model of work engagement. Career Dev. Int. 13 (3), 209–223. Bakker, A.B., Leiter, M.P., 2010. Where to go from here: integration and future research on work engagement. In: Bakker, A.B., Leiter, M.P. (Eds.), Work Engagement: A Handbook of Essential Theory and Research. Psychology Press, New York, NY, pp. 181–196. Bandalos, D.L., Finney, S.J., 2001. Item parceling issues in structural equation modeling. In: Marcoulides, G.A., Schumacker, R.E. (Eds.), Advanced Structural

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

25

Equation Modeling: New Developments and Techniques. Lawrence Erlbaum Associates, Mahwah, NJ, pp. 269–296. Baron, R.M., Kenny, D.A., 1986. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J. Pers. Soc. Psychol. 51 (6), 1173–1182. Baumruk, R., 2004. The missing link: the role of employee engagement in business success. Workspan 47 (11), 48–52. Bollen, K.A., 1989. Structural Equations with Latent Variables. John Wiley & Sons, New York, pp. NY–$9. Bollen, K.A., Stine, R., 1990. Direct and indirect effects: classical and bootstrap estimates of variability. Sociol. Methodol. 20, 115–140. Buckley, M.R., Cote, J.A., Comstock, S.M., 1990. Measurement errors in behavioral sciences: the case of personality/attitude research. Educ. Psychol. Meas. 50 (3),

447–474.

Burke, R.J., Koyuncu, M., Jing, W., Fiksenbaum, L., 2009. Work engagement among hotel managers in Beijing, China: potential antecedents and consequences. Tourism Rev. 64 (3), 4–18. Byrne, B.M., 2001. Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming. Lawrence Erlbaum Associates, London. Campbell, D.T., Fiske, D.W., 1959. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol. Bull. 56 (2), 81–105. Cha, J., Kim, S., Cichy, R.F., Kim, M., Tkach, J.L., 2013. General managers’ and chief officers’ evaluations of private club boards of directors. Int. J. Hospitality Manage. 32 (March), 245–253. Chen, C.-Y., Yen, C.-H., Tsai, F.C., 2014. Job crafting and job engagement: the mediating role of person-job fit. Int. J. Hospitality Manage. 37 (February),

21–28.

Chi, C.G., Gursoy, D., 2009. Employee satisfaction, customer satisfaction, and financial performance: an empirical examination. Int. J. Hospitality Manage. 28 (2), 245–253. Christian, M.S., Garza, A.S., Slaughter, J.E., 2011. Work engagement: a quantitative review and test of its relations with task and contextual performance. Personnel Psychol. 64 (1), 89–136.

Churchill Jr., G.A., 1979. A paradigm for developing better measures of marketing constructs. J. Mark. Res. 16 (1), 64–73. Clapp-Smith, R., Vogelgesang, G.R., Avey, J.B., 2009. Authentic leadership and positive psychological capital: the mediating role of trust at the group level of analysis. J. Leadersh. Organ. Studies 15 (3), 227–240. Culbertson, S.S., Fullager, C.J., Mills, M.J., 2010. Feeling good and doing great: the relationship between psychological capital and well-being. J. Occup. Health Psychol. 15 (4), 421–433. Field, A., Hole, G.J., 2003. How to Design and Report Experiments. Sage Publications, London. Fornell, C., Larcker, D., 1981. Evaluating structural equation models with unobservable and measurement error. J. Mark. Res. 18 (1), 39–50. Gooty, J., Gavin, M., Johnson, P.D., Frazier, M.L., Snow, D.B., 2009. In the eyes of the beholder: transformational leadership, positive psychological capital, and performance. J. Leadersh. Organ. Stud. 15 (4), 353–367. Grandey, A.A., Cropanzano, R., 1999. The conservation of resources model applied

to work-family conflict and strain. J. Vocational Behav. 54 (2), 350–370.

Grönroos, C., 2000. Service Management and Marketing: A Customer Relationship Approach, 2nd ed. John Wiley & Sons, Chichester.

Hair Jr., J.F., Black, W.C., Babin, B.J., Anderson, R.E., 2010. Multivariate Data Analysis:

A global Perspective, 7th ed. Pearson Prentice Hall, Upper Saddle River, NJ.

Hartline, M.D., Ferrell, O.C., 1996. The management of customer-contact service employees: An empirical investigation. J. Mark. 60 (4), 52–70. Hall, R.J., Snell, A.F., Foust, M.S., 1999. Item parceling strategies in SEM:

investigating the subtle effects of unmodeled secondary constructs. Organ. Res. Methods 2 (3), 233–256. Harter, J.K., Schmidt, F.L., Hayes, T.L., 2002. Business-unit level relationship between employee satisfaction, employee engagement, and business outcomes: a meta-analysis. J. Appl. Psychol. 87 (2), 268–279. Hobfoll, S.E., 1989. Conservation of resources: a new attempt at conceptualizing stress. Am. Psychol. 44 (3), 513–524. Hobfoll, S.E., 2001. The influence of culture, community, and the nested-self in the stress process: advancing conservation of resources theory. Appl. Psychol. Int. Rev. 50 (3), 337–421. Hon, A.H.Y., 2011. Enhancing employee creativity in the Chinese context: the mediating role of employee self-concordance. Int. J. Hospitality Manage. 30 (2),

375–384.

Jafri, H., 2012. Psychological capital and innovative behavior: an empirical study on apparel fashion industry. J. Contemp. Manage. Res. 6 (1), 42–52. Johnson, T., Fendrich, M., 2005. Modeling sources of self-report bias in a survey of drug use epidemiology. Ann. Epidemiol. 15 (5), 381–389. Kahn, W.A., 1990. Psychological conditions of personal engagement and disengagement at work. Acad. Manage. J. 33 (4), 692–724. Karatepe, O.M., 2011. Procedural justice, work engagement, and job outcomes:

evidence from Nigeria. J. Hospitality Mark. Manage. 20 (8), 855–878. Karatepe, O.M., 2013a. High-performance work practices and hotel employee performance: the mediation of work engagement. Int. J. Hospitality Manage. 32 (1), 132–140. Karatepe, O.M., 2013b. Perceptions of organizational politics and hotel employee outcomes: the mediating role of work engagement. Int. J. Contemp. Hospitality Manage. 25 (1), 82–104.

Karatepe, O.M., 2014. Hope, work engagement, and organizationally valued performance outcomes: an empirical study in the hotel industry. J. Hospitality Mark. Manage. 23 (6), 678–698. Karatepe, O.M., Beirami, E., Bouzari, M., Safavi, H.P., 2014. Does work engagement mediate the effects of challenge stressors on job outcomes? Evidence from the hotel industry. Int. J. Hospitality Manage. 36 (January), 14–22. Karatepe, O.M., Demir, E., 2014. Linking core-self-evaluations and work engagement to work-family facilitation: a study in the hotel industry. Int. J. Contemp. Hospitality Manage. 26 (2), 307–323.

Karatepe, O.M., Karadas, G., 2014. The effect of psychological capital on conflicts in the work-family interface, turnover and absence intentions. Int. J. Hospitality Manage. 43 (October), 132–143. Karatepe, O.M., Karadas, G., Azar, A.K., Naderiadib, N., 2013. Does work engagement mediate the effect of polychronicity on performance outcomes? A study in the hospitality industry in Northern Cyprus. J. Hum. Resour. Hospitality Tourism 12 (1), 52–70. Karatepe, O.M., Ngeche, R.N., 2012. Does job embeddedness mediate the effect of work engagement on job outcomes? A study of hotel employees in Cameroon. J. Hospitality Mark. Manage. 21 (4), 440–461. Karatepe, O.M., Olugbade, O.A., 2009. The effects of job and personal resources on hotel employees’ work engagement. Int. J. Hospitality Manage. 28 (4), 504–512. Karatepe, O.M., Uludag, O., Menevis, I., Hadzimehmedagic, L., Baddar, L., 2006. The effects of selected individual characteristics on front-line employee performance and job satisfaction. Tourism Manage. 27 (4), 547–560. Kim, H.J., Shin, K.H., Swanger, N., 2009. Burnout and engagement: a comparative analysis using the Big Five personality dimensions. Int. J. Hospitality Manage.

28 (1), 96–104.

Kim, T.T., Lee, G., 2013. Hospitality employee knowledge-sharing behaviors in the relationship between goal orientations and service innovative behavior. Int. J. Hospitality Manage. 34 (September), 324–337. King, M.F., Bruner, G.C., 2000. Social desirability bias: a neglected aspect of validity testing. Psychol. Marke. 17 (2), 79–103. Kline, R.B., 2010. Principles and Practice of Structural Equation Modeling, 3rd ed. Guilford Press, New York, NY. Larson, M., Luthans, F., 2006. Potential added value of psychological capital in predicting work attitudes. J. Leadersh. Organ. Stud. 13 (2), 75–92. Leiter, M.P., Bakker, A.B., 2010. Work engagement: introduction. In: Bakker, A.B., Leiter, M.P. (Eds.), Work Engagement: A Handbook of Essential Theory and Research. Psychology Press, New York, NY, pp. 1–9. Leung, A.S.M., Wu, L.Z., Chen, Y.Y., Young, M.N., 2011. The impact of workplace ostracism in service organizations. Int. J. Hospitality Manage. 30 (4), 836–844. Li, X.B., Sandersb, K., Frenkelc, S., 2012. How leader–member exchange, work

engagement and HRM consistency explain Chinese luxury hotel employees’ job performance. Int. J. Hospitality Manage. 31 (4), 1059–1066.

Llorens, S., Schaufeli, W., Bakker, A., Salanova, M., 2007. Does a positive gain spiral of resources, efficacy beliefs and engagement exists? Comput. Hum. Relations

23 (1), 825–841.

Locke, E.A., 1976. The nature and causes of job satisfaction. In: Dunnette, M.D. (Ed.), Handbook of Industrial and Organizational Psychology. Rand McNally College Publishing Company, Chicago, IL, pp. 1297–1349. Luthans, F., Avolio, B.J., Avey, J.B., Norman, S.M., 2007a. Positive psychological capital: measurement and relationship with performance and satisfaction. Personnel Psychol. 60 (3), 541–572. Luthans, F., Avolio, B.J., Walumbwa, F.O., Li, W.X., 2005. The psychological capital of Chinese workers: exploring the relationship with performance. Manage. Organ. Rev. 1 (2), 249–271. Luthans, F., Jensen, S.M., 2005. The linkage between psychological capital and commitment to organizational mission: a study of nurses. J. Nurs. Admin. 35 (6), 304–310. Luthans, F., Norman, S.M., Avolio, B.J., Avey, J.B., 2008. The mediating role of psychological capital in the supportive organizational climate: employee performance relationship. J. Organ. Behav. 29 (2), 214–238. Luthans, F., Youssef, C.M., 2007. Emerging positive organizational behavior. J. Manage. 33 (3), 321–349. Luthans, F., Youssef, C.M., Avolio, B.J., 2007b. Psychological Capital: Developing the Human Competitive Edge. Oxford University Press, Oxford, UK. MacKinnon, D.P., Lockwood, C.M., Hoffman, J.M., West, S.G., Sheets, V., 2002. A comparison of methods to test mediation and other intervening variable effects. Psychol. Methods 7 (1), 83–104. Maltz, A.C., Shenhar, A.J., Reilly, R.R., 2003. Beyond the balanced scorecard: refining the search for organizational success measures. Long Range Plann. 36 (2),

187–204.

Maslach, C., Leiter, M.P., 1997. The Truth about Burnout: How Organizations Cause Personal Stress and What to Do about It. Jossey-Bass, San Francisco, CA. Mattila, A.S., Enz, C.A., 2002. The role of emotions in service encounters. J. Service Res. 4 (4), 268–277. Mossholder, K.W., Bennett, N., Kemery, E.R., Wesolowski, M.A., 1998. Relationships between bases of power and work reactions: the meditational role of procedural justice. J. Manage. 24 (4), 533–552. Nelson, D., Cooper, C.L., 2007. Positive Organizational Behavior: Accentuating the Positive at Work. Sage Publications, Thousand Oaks, CA. Netemeyer, R.G., Durvasula, S., Lichtenstein, D.R., 1991. A cross-national assessment of the reliability and validity of the CETSCALE. J. Mark. Res. 28 (3),

320–327.

26

S. Paek et al. / International Journal of Hospitality Management 50 (2015) 9–26

Norman, S.M., Avey, J.B., Nimnicht, J.L., Pigeon, N.G., 2010. The interactive effects of psychological capital and organizational identity on employee organizational citizenship and deviance behaviors. J. Leadersh. Organ. Stud. 17 (4), 380–391. Nunnally, J.C., 1978. Psychometric Theory, 2n ed. McGraw-Hill, New York, NY. Onsøyen, L.E., Mykletun, R.J., Steiro, T.J., 2009. Silenced and invisible: the work-experience of room-attendants in Norwegian hotels. Scand. J. Hospitality Tourism 9 (1), 81–102. Park, J., Gursoy, D., 2012. Generation effects on work engagement among U.S. hotel employees. Int. J. Hospitality Manage. 31 (4), 1195–1202. Peterson, C., Seligman, M., 2004. Character Strengths and Virtues: A Handbook and Classification. Oxford University Press, New York, NY. Pienaar, J., Willemse, S.A., 2008. Burnout, engagement, coping and general health of service employees in the hospitality industry. Tourism Manage. 29 (6),

1053–1063.

Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol. 88 (5), 879–903. Podsakoff, P.M., Organ, D.W., 1986. Self-reports in organizational research:

problems and prospects. J. Manage. 12 (4), 531–544. Preacher, K.J., Hayes, A.F., 2004. SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behav. Res. Methods Instruments Comput. 36 (4), 717–731. Rego, A., Sousa, F., Marques, C., Cunha, M.P., 2012. Authentic leadership promoting employees’ psychological capital and creativity. J. Bus. Res. 65 (3), 429–437. Rust, R.T., Lee, C., Valente, E., 1995. Comparing covariance structure models: a general methodology. Int. J. Res. Mark. 12 (4), 279–291. Salanova, M., Agut, S., Peiro, J.M., 2005. Linking organizational resources and work engagement to employee performance and customer loyalty: the mediation of service climate. J. Appl. Psychol. 90 (6), 1217–1227. Schaufeli, W.B., Bakker, A.B., 2004. Job demands, job resources, and their relationship with burnout and engagement: a multi-sample study. J. Organ. Behav. 25 (3), 293–315. Schaufeli, W.B., Bakker, A.B., Salanova, M., 2006. The measurement of work engagement with a short questionnaire: a cross-national study. Educ. Psychol. Meas. 66 (4), 701–716. Schaufeli, W.B., Martinez, I.M., Pinto, A.M., Salanova, M., Bakker, A.B., 2002. Burnout and engagement in university students: a cross-national study. J. Cross-Cultural Psychol. 33 (5), 464–481.

Shrout, P.E., Bolger, N., 2002. Mediation in experimental and nonexperimental studies: new procedures and recommendations. Psychol. Methods 7 (4),

422–445.

Slåtten, T., Mehmetoglu, M., 2011. Antecedents and effects of engaged front-line employees: a study from the hospitality industry. Manag. Service Qual. 21 (1),

88–107.

Sobel, M.E., 1982. Asymptotic confidence intervals for indirect effects in structural equation models. In: Leinhardt, S. (Ed.), Sociological Methodology, 13. Jossey-Bass, San Francisco, CA, pp. 290–312. Stajkovic, A.D., Luthans, F., 1998. Self-efficacy and work-related performance: a meta analysis. Psychol. Bull. 124 (2), 240–261. Sulea, C., Virga, D., Maricutoiu, L.P., Schaufeli, W., Dumitru, C.Z., Savai, F.A., 2012. Work engagement as mediator between job characteristics and positive and negative extra-role behaviors. Career Dev. Int. 17 (3), 177–207. Sweetman, D., Luthans, F., 2010. The power of positive psychology: psychological capital and work engagement. In: Bakker, A.B., Leiter, M.P. (Eds.), Work Engagement: A Handbook of Essential Theory and Research. Psychology Press, New York, NY, pp. 54–68. Wang, H., Sui, Y., Luthans, F., Wang, D., Wu, Y., 2012. Impact of authentic leadership on performance: role of followers’ positive psychological capital and relational processes. J. Organ. Behav. 35 (1), 5–21. Williams, L., O’Boyle, E., 2008. Measurement models for linking latent variables and indicators: a review human resource management research using parcels. Hum. Resour. Manage. Rev. 18 (4), 233–242. Xanthopoulou, D., Bakker, A.B., Demerouti, E., Schaufeli, W.B., 2009. Reciprocal relationships between job resources, personal resources, and work engagement. J. Vocational Behav. 74 (3), 235–244. Yalabik, Z.Y., Popaitoon, P., Chowne, J.A., Rayton, B.A., 2013. Work engagement as a mediator between employee attitudes and outcomes. Int. J. Hum. Resour. Manage. 24 (14), 1799–2823. Yang, F.X., Lau, V.M.-C., 2015. Does workplace guanxi to matter hotel career success? Int. J. Hospitality Manage. 47 (May), 43–53. Yeh, C.M., 2013. Tourism involvement, work engagement and job satisfaction among front-line hotel employees. Ann. Tourism Res. 42, 214–239.