The Relationship Between Cross-Cultural Psychological Capital and Organizational Commitment.

Chapter 1: Introduction to the Study

Introduction

Advancements in communication and transportation technologies have

contributed to increased globalization worldwide, resulting in many organizations

becoming more culturally diverse than ever before (Christensen & Kowalczyk, 2017;

Wood & Wilberger, 2015). The advantages of a diverse workforce, such as improved

customer focus and satisfaction, a broader skills base (O’Neill, 2016), and higher

motivation (Kotze & Massyn, 2019), are well-known. Employees’ ability to interact

within these multicultural environments can enhance or detract from the positive

organizational outcomes their organizations are hoping to achieve (Adler & Aycan,

2018). Researchers have suggested that in multicultural organizations, employees may

struggle with cultural competency. The adverse effects of these struggles impact not only

the intercultural interactions between employees and positive organizational outcomes

(Adler & Aycan, 2018) but also the psychological health and well-being of the employees

within the organization (Blanchet-Garneau & Pepin, 2015; Kotze & Massyn, 2019).

In their annual review on psychological capital (PsyCap), Luthans and Youssef-

Morgan (2017) suggested that while there is a modest body of research on PsyCap,

further research is required to understand better how PsyCap functions under context-

specific conditions. In making this suggestion, Luthans and Youssef-Morgan highlighted

cross-cultural and culturally diverse contexts as a research area needing development.

Consequently, Dollwet and Reichard (2014) conceptualized cross-cultural PsyCap as a

distinct collection of personal psychological resources which allows individuals to

positively adjust to cross-cultural interactions in culturally diverse workplaces (Kotze &

Massyn, 2019; Maslakci & Sesen, 2019; Yunlu & Clapp-Smith, 2014). Studies have

shown that high cross-cultural PsyCap is associated with a higher likelihood of positive

inter-cultural interactions (Dollwet & Reichard, 2014) and specific positive

organizational outcomes such as employee engagement (Kotze & Massyn, 2019).

However, there is limited knowledge of how positive interactions impact other positive

organizational outcomes such as organizational commitment (OC).

Researchers have identified OC as a positive organizational outcome crucial to

employees’ performance and turnover intentions (Allen & Meyer, 1996; Meyer et al.,

2002; Meyer et al., 2004). While the benefits of improved performance to both

employees and the organization are apparent, the benefits of reduced turnover intentions

may be just as significant. Researchers have been aware of numerous consequences of

turnover for many years, including increased costs related to recruitment, training, and

development, disruption in operations, and the demoralization of those who remain

(Abbasi & Hollman, 2000; Staw, 1980). However, research conducted using nurses in

health care organizations has identified that turnover increased costs related to

recruitment and replacement (Halter et al., 2017) decreased productivity, increased the

pressure on those that remained, and led to reduced patient outcomes (Dewanto &

Wardhani, 2017; Hayes et al., 2006).

This study examines the important relationship between cross-cultural PsyCap

and OC for employees in a multicultural workplace such as a health care organization.

The study will add to the body of knowledge highlighting the relationship between cross-

cultural PsyCap and positive organizational outcomes, further contributing to the

development of the construct while also providing support for achieving social change

through improving cultural competence within multicultural workplaces.

Background

Globalization has changed organizations and the cultural demographics

represented in workforces worldwide (Wood & Wilberger, 2015). More than ever before,

changes in the cultural demographics of organizations’ workforces have created scenarios

where employees must develop and leverage new skills to successfully work with co-

workers from different cultures who may speak other languages and have different beliefs

(Dollwet & Reichard, 2014). While research has shown numerous advantages to

culturally diverse workforces, including increased employee motivation (Kotze &

Massyn, 2019) and increased job satisfaction (Bergheim et al., 2015), the psychological

resources used to function effectively across different cultures to achieve those results are

of vital importance. This study will address a gap in the literature about cross-cultural

PsyCap and the predictive relationship between cross-cultural PsyCap and OC in

multicultural organizations.

Psychological Capital

PsyCap is a psychological state of development based on the personal

psychological resources of hope, efficacy, resilience, and optimism (Luthans & Youssef-

Morgan, 2017), commonly referred to by the acronym HERO. PsyCap, as a construct,

was developed under the broader movements of positive organizational behavior (POB),

positive organizational scholarship (POS), and most broadly, positive psychology (PP).

These areas of psychology, along with PsyCap as a construct, focus primarily on

individuals’ positive experiences and traits rather than the maladaptive behaviors

individuals engage in (Seligman, 2019). However, to be included within these domains of

psychology, PsyCap must also display several other characteristics, including being state-

like and context-specific (Luthans & Youssef-Morgan, 2017). State-like, in this context,

means that PsyCap is somewhat flexible, and employees can develop their PsyCap

through external interventions such as training and development sessions (Dollwet &

Reichard, 2014; Luthans & Youssef-Morgan, 2017). Context or domain-specificity

implies that PsyCap can change in different contexts, such as a workplace or educational

setting, whereas a person may have high levels of HERO in one, and they may have

much lower levels in the other (Dollwet & Reichard, 2014; Luthans & Youssef-Morgan,

2017). PsyCap is also a higher-order construct as it is a latent variable that is observable

through the combination of the HERO psychological resources. As such, researchers

often measure PsyCap through questionnaires that have individual scales reflecting each

of the individual HERO components. PsyCap studies have found positive relationships

between PsyCap and job satisfaction (Badran & Youssef-Morgan, 2015; Bergheim et al.,

2015) while also identifying correlations between positive safety perceptions and high

levels of PsyCap (Bergheim et al., 2015). Other research has suggested that employees

who experienced positive PsyCap were more likely to engage in organizational

citizenship behaviors, and team leaders’ learning values and optimism strengthened the

relationship (Bogler and Somech, 2019).

Additionally, Firestone and Anngela-Cole (2016) found evidence suggesting a

relationship between certain quality of life factors external to an organization and

PsyCap. The researchers found that the psychological resources that combine to form

PsyCap (i.e., HERO) could be measured reliably in non-profit organizations. Results

were consistent with previous data specific to for-profit organizations. However, not all

studies on PsyCap have achieved positive results; Idris and Manganaro (2017) could not

find any significant relationships between PsyCap, job satisfaction, and OC for a specific

population of Saudi Arabian petrochemical workers. These results aside, the

preponderance of studies have shown that higher levels of PsyCap are generally

associated with high levels of positive outcomes for employees and organizations.

Cross-Cultural Psychological Capital

Whereas PsyCap is a psychological state of development based on HERO

(Luthans & Youssef-Morgan, 2017), cross-cultural PsyCap extends the construct into the

domain of intercultural interactions within a multicultural workplace (Dollwet &

Reichard, 2014). Recalling the domain-specificity of the PsyCap construct, cross-cultural

PsyCap focuses on the HERO psychological resources in terms of interactions that occur

between individuals from different cultures within the workplace (Dollwet & Reichard,

2014). It is less concerned with how individuals feel about their ability to complete their

work successfully and more concerned with their perceptions and feelings towards their

ability to successfully navigate and interact among diverse groups of individuals within a

work environment. Similar to PsyCap, cross-cultural PsyCap is a latent variable

observable through the cross-cultural HERO components. However, the differentiation

between the HERO and the cross-cultural HERO scales is essential as the scales measure

different resources and may provide profoundly different results. While the workplace

hope scale intends to measure an employee’s ability to set and achieve goals related to

their work, the cross-cultural hope scale changes the focus to the measure from work

activities to intercultural interactions (Dollwet & Reichard, 2014). While the

psychological resource is similar, the target of the resource is vastly different. To this

end, Dollwet and Reichard (2014) adapted Luthans et al.’s (2007a) PsyCap questionnaire

to measure cross-cultural PsyCap related to intercultural interactions in organizations

comprised of a culturally diverse group of employees. The researchers analyzed the tools

for psychometric properties, and the result indicated that the tools measuring cross-

cultural PsyCap were measuring a distinct construct. Other cross-cultural PsyCap studies

have found cross-cultural PsyCap to be an indicator of cultural competence and suggested

increased PsyCap/cultural competence resulted in higher levels of employee well-being

in a group of South African employees (Kotze & Massyn, 2019). A study using a sample

of hospitality employees in Northern Cyprus further identified that cross-cultural PsyCap

assisted with intercultural competencies and that cross-cultural PsyCap mediated the

relationship between multicultural personality traits and perceived service quality

(Maslakci & Sesen, 2019). Finally, Yunlu and Clapp-Smith (2014) also adapted the

PsyCap questionnaire. Their research found that cross-cultural PsyCap was strongly

related to motivational cultural intelligence and metacognitive awareness; Yunlu and

Clapp-Smith also confirmed while the three constructs are conceptually similar, they are

distinctly separate. These findings support the results found in the earlier studies on

PsyCap in terms of the relationship with positive organizational outcomes; however, the

domain of intercultural interactions in multicultural workplaces has only received

minimal examination, and gaps remain in current literature. In particular, gaps in the

literature remain in terms of intercultural interactions related to other positive

organizational outcomes such as OC.

Organizational Commitment

Research on OC has continued to grow exponentially since the mid-1970s

(Mowday, 1998); however, in terms of construct popularity, few OC constructs could

compare to the popularity achieved by Meyer and Allen’s (1991) three-component model

of OC. Initially, Allen and Meyer defined OC as “a psychological link between the

employee and his or her organization that makes it less likely that the employee will

voluntarily leave the organization” (Allen & Meyer, 1996, p. 252). Subsequently, Meyer

and Herscovitch (2001) developed a more generalized definition of OC by developing a

general workplace commitment model. Meyer and Herscovitch proposed a new definition

where “commitment is a force that binds an individual to a course of action of relevance

to one or more targets” (p. 301). This new definition highlights that the

individual/employee may be bound to something else such as their supervisor or a

specific project rather than the organization. The literature on the three-component model

of commitment separates the various concepts of OC into three main components,

affective commitment (AC), normative commitment (NC), and continuance commitment

(CC; Meyer & Allen, 1991). While AC is understood to describe an employee’s

emotional attachment towards an organization (Allen & Meyer, 1996), other

characteristics include a desire to pursue an action relevant to a target to which the

employee is emotionally attached (Meyer & Herscovitch, 2001). NC is commonly

understood to describe a sense of obligation towards an organization (Allen & Meyer,

1996); however, NC is also used to indicate a feeling that an individual is obligated to

engage in a particular action or behavior relevant to a target (Meyer & Herscovitch,

2001). Finally, researchers often define CC as commitment based on the cost of

alternatives (Allen & Meyer, 1996); however, they may also define CC in terms of the

costs associated with ceasing a behavior or action relevant to a target (Meyer &

Herscovitch, 2001). Studies on AC have shown a robust correlation with other positive

organizational outcomes (Meyer et al., 2002; Meyer et al., 2004), including a predictive

relationship with turnover. However, a few studies have shown in certain cultural

contexts (i.e., collectivist cultures) that NC is a stronger predictor of turnover than AC

(Chang et al., 2007; Vandenberghe, 2003). Similar to NC, some studies have shown

cultural context to have an impact on CC. Certain cultures will seek to remove ambiguity

by establishing precise alternatives and, as such, often exhibit much higher levels of CC

as they perceive higher costs and fewer viable alternatives (Chang et al., 2007). While

studies have found there to be a net positive correlation between PsyCap and OC

(Hussain & Nawaz, 2019; Sen et al., 2017; Surucu et al., 2020; Yildiz, 2018), there

remains a gap when it comes to the relationship between domain-specific PsyCap (i.e.,

cross-cultural PsyCap) and OC. It remains to be seen whether a similar relationship will

endure in a different context. This study will examine that gap and highlight the specific

relationships between the psychological resources in cross-cultural PsyCap and each of

the three components under OC.

Problem Statement

Organizations continue to pursue opportunities to establish a competitive

advantage within a globalized environment (Christensen & Kowalczyk, 2017; Wood &

Wilberger, 2015). The development of workplace PsyCap in employees, including

context-specific forms, such as cross-cultural PsyCap, have been suggested as avenues

through which organizations can build and maintain a competitive advantage in globally

competitive marketplaces (Luthans & Youssef-Morgan, 2017). Furthermore, the

collection of cross-cultural psychological resources that form cross-cultural PsyCap may

indicate individual employees’ ability to develop the cross-cultural competencies needed

to succeed within culturally diverse work contexts (Dollwet & Reichard, 2014; Kotze &

Massyn, 2019). Emerging research specific to cross-cultural PsyCap has provided some

encouraging results about the positive relationships between cross-cultural PsyCap and

specific positive organizational outcomes such as employee well-being (Kotze &

Massyn, 2019) and service quality (Maslakci & Sesen, 2019). However, many other

positive outcomes, such as OC, have yet to be explored (Kotze & Massyn, 2019;

Maslakci & Sesen, 2019).

The specific problem is that as workplaces become more culturally diverse, it is

essential to explore how organizations can leverage positive psychological resources to

enhance AC, NC, and CC. Yet, the nature of cross-cultural PsyCap’s influence on OC is

unclear. Results from existing research indicate that cross-cultural PsyCap has a positive

effect on workplace well-being and a negative impact on burnout (Kotze & Massyn,

2019). However, to continue developing the intercultural interaction domain of the

construct, further exploration into cross-cultural PsyCap’s influence on positive

organizational outcomes (Kotze & Massyn, 2019) and generalization of the construct in

different industries (Maslakci & Sesen, 2019) must occur. Although globalization and

social equality continue to influence and increase cultural diversity within organizations,

employees’ cultural competence has not increased at the same level. Suppose

organizations do not acknowledge and address this disparity. In that case, they will not be

able to leverage the advantages of a culturally diverse workforce, and employees may

experience adverse psychological effects, resulting in reduced organizational success.

Purpose of the Study

The purpose of this quantitative correlational research is to explore the nature of

the relationship between cross-cultural PsyCap and OC for employees at a health care

organization in Canada. This study defines cross-cultural PsyCap as a context-specific

form of PsyCap consisting of several specific psychological resources (Kotze & Massyn,

2019). The personal psychological resources included in this definition are cross-cultural

hope, cross-cultural self-efficacy, cross-cultural resilience, and cross-cultural optimism

(Kotze & Massyn, 2019). This study defines OC as a psychological attachment between

the employee and the organization consisting of three sub-dimensions including AC

(employees’ emotional attachment), NC (employees’ sense of obligation), and CC

(employees’ evaluation of loss if they were to leave; Peng et al., 2013). The personal

psychological resources that make up cross-cultural PsyCap are the latent variables

contributing to cross-cultural PsyCap, the predictor variable. Similarly, AC, NC, and CC

are the criterion variables contributing to overall OC.

The results of this study may help advance current theory on cross-cultural

PsyCap. The results of this study may also contribute to the development of training and

development interventions within Canadian health care organizations through

highlighting the value, in terms of positive organizational outcomes, of developing and

supporting employees’ intercultural interaction skills.

Research Questions and Hypotheses

The following research questions and hypotheses were used in the current study

and are visually depicted in Figures 1, 2, 3, and 4.

RQ1 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the AC component of OC in employees at a Canadian health care

organization?

H01 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

Ha1 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

RQ2 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the NC component of OC in employees at a Canadian health care

organization?

H02 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

Ha2 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

RQ3 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the CC component of OC in employees at a Canadian health care

organization?

H03 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

Ha3 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

RQ4 – Quantitative: Does Canadian health care organization employees’ type of

employment influence the relationship between their cross-cultural PsyCap and OC?

H04 – Canadian health care organization employees’ type of employment does not

influence the relationship between their cross-cultural PsyCap and OC.

Ha4 – Canadian health care organization employees’ type of employment

influences the relationship between their cross-cultural PsyCap and OC.

Figure 1

Research Question 1

Figure 2

Research Question 2

Figure 3

Research Question 3

Figure 4

Research Question 4

Theoretical Framework

The theoretical base for this study is Hobfoll’s (1989) theory of the conservation

of resources (COR). Hobfoll’s COR theory is a stress and motivation theory that

identifies that employees who have or gain resources, including psychological resources,

will pursue more resources to maintain their resources and prevent against future loss

(Hobfoll, 2011). The theory of COR is rooted in four principles: 1) resource loss is

disproportionately more impactful than resource gain, 2) resources must be invested to

prevent loss, to maintain, or to gain other resources, 3) when resources are low, or

resource loss is high, resource gain becomes more significant and valuable, and 4)

individuals will act defensively or in desperation to protect what they have when

resources are scarce (Hobfoll et al., 2018). These principles assist in understanding the

motivations behind an employee’s response to stress and challenges in the workplace

(Hobfoll, 2011). For example, if an employee’s health, well-being, family, or sense of

self are at risk, the behaviors the employee engages in to respond to the threat are

understandable in the context of the four principles of COR theory.

Liu (2014) described cross-cultural PsyCap as “a type of cross-cultural dynamic

competency in the family of personal resources” which “can be regarded as a cross-

cultural personal resource in Hobfoll’s COR theory’s resource family” (p. 83). In

practical terms, this means that individuals who develop (or already have) the personal

psychological resources that make up cross-cultural PsyCap will actively seek to

maintain or increase those resources through increased competence (Kotze & Massyn,

2019). To grow, maintain, or even protect against loss, the individual will have to invest a

certain amount of their already existing resources. Furthermore, Hobfoll (2011) suggested

individuals who have significant psychological resources may exhibit more resilience in

stressful situations. This suggestion aligns with the conceptualization of resilience in

cross-cultural PsyCap; individuals are more likely to achieve better results when

confident in their ability to overcome communication issues and other obstacles when

interacting with individuals from different cultures.

Consistent with COR theory, OC is a psychological state in which employees

evaluate their relationship with the organization and subsequently decide whether they

will stay with the organization or leave (Meyer & Allen, 1991). As such, any loss in

resources that contribute to a positive evaluation of the relationship would be

disproportionately more impactful than any increase (i.e., the primacy of loss).

Additionally, Hobfoll’s (2011) conceptualization of resource caravan passageways

suggests that the environmental conditions in which employees operate (i.e., physical

conditions, organizational culture) either enhance or impede the process of resource loss,

maintenance, or development. In terms of OC, employees will strive to sustain or

increase resources, particularly those that contribute to a positive evaluation of their

relationship with the organization. The improved assessment of the employee-

organization relationship reflects an increase in one or more components of commitment.

However, organizational factors external to the employee influence an employee’s ability

to gain, maintain, or even lose resources, resulting in either stress or motivation, which

are central to COR theory. Previous research has shown positive correlations between

specific personal psychological resources such as optimism, hope, and self-efficacy and

the components of OC (Durukan Kose et al., 2018; Yildiz, 2018). However, such results

have not been generalized to the domain of intercultural interactions.

Nature of the Study

The nature of this study was quantitative research using a cross-sectional,

correlational design and self-reported data collected through surveys. This study was

designed to focus on the nature of the relationship between cross-cultural PsyCap and

OC. As such, the study was well-suited to a cross-sectional design using an online survey

method (Creswell & Creswell, 2018). Furthermore, as Groves et al. (2009) discussed,

while there is no one ideal method for every situation, some research methods more than

others serve better for certain research attributes and scenarios. For example, health care

organizations may have workforces in the tens of thousands of employees (or more). As

such, an online survey tool is a cost-effective method of reaching out to a large

population of employees who may work at numerous different facilities in several other

geographic locations.

The cross-sectional design also captured the participants’ self-reported data at the

point in time when they were answering the survey. Capturing the data in this way was

important as cross-cultural PsyCap is a state-like resource (Dollwet & Reichard, 2014).

Although state-like resources are more stable than pure states, state-like resources are still

malleable and can change due to external influences, particularly over extended periods

(Luthans & Youssef-Morgan, 2017). In a longitudinal design, factors such as changes in

leadership, organizational culture, or even employment level (i.e., front-line, supervisory,

etc.) can influence a malleable state-like resource such as cross-cultural PsyCap, which

may confound the results. Additionally, prior literature (e.g., Kotze & Massyn, 2019;

Maslakci & Sesen, 2019) has established that a cross-sectional design is acceptable for

research on this topic, and more analysis of this type is needed to inform the literature in

this area better.

This research was well-suited to a quantitative approach. The survey collected

self-reported data on cross-cultural PsyCap and OC from a population of employees at a

large health care organization in Canada. The intention was that the representative sample

would consist of three different employment groups proportionate to the size of the

existing groups within the population. The predictor variable in this research, cross-

cultural PsyCap, reflects the personal cross-cultural psychological resources (i.e., cross-

cultural HERO). In contrast, OC is a reflective variable consisting of the three

components of commitment (i.e., affective, normative, and continuance), which are the

criterion variables. Cross-cultural PsyCap was measured using Dollwet and Reichard’s

(2014) 20-item cross-cultural PsyCap scale. OC was measured using Meyer and

Herscovitch’s (2001) revised version of Meyer and Allen’s (1991) three-component

model of employee commitment scale; the revised version uses 18-items across three

scales to measure the three components of OC.

Definition of Terms

The following terms are operationalized for the study as follows:

Affective commitment: Affective commitment describes an employee’s

“identification with, involvement in, and emotional attachment to the organization”

(Allen & Meyer, 1996, p. 253).

Continuance commitment: Continuance commitment describes an employee’s

“commitment based on the employee’s recognition of the costs associated with leaving

the organization” (Allen & Meyer, 1996, p. 253).

Cross-cultural hope: Cross-cultural hope refers to a personal psychological

resource focused on “pursuing and meeting goals related to working with people from

different cultures” (Dollwet & Reichard, 2014, p. 1672).

Cross-cultural optimism: Cross-cultural optimism is a personal psychological

resource focused on being able to “remain positive and motivated during cross-cultural

interaction” (Kotze & Massyn, 2019, p. 2)

Cross-cultural psychological capital: Cross-cultural psychological capital

describes a state-like, context-specific construct where the psychological resources of

cross-cultural hope, cross-cultural self-efficacy, cross-cultural resilience, and cross-

cultural optimism are applied directly to inter-cultural interactions in the workplace

(Dollwet & Reichard, 2014).

Cross-cultural resilience: Cross-cultural resilience refers to a personal

psychological resource focused on employees being “able to keep up their performance

despite obstacles, such as language difficulties, cross-cultural conflict or other cross-

cultural issues” (Kotze & Massyn, 2019, p. 2).

Cross-cultural self-efficacy: Cross-cultural self-efficacy is a personal

psychological resource focused on having confidence in one’s own ability to interact and

communicate with employees from different cultural groups (Dollwet & Reichard, 2014;

Kotze & Massyn, 2019).

Cultural competence: Cultural competence is a concept focused on understanding

different cultures, enabling individuals to effectively communicate with others to achieve

successful inter-cultural interactions (Kotze & Massyn, 2019).

Normative commitment: Normative commitment describes an employee’s

“commitment based on a sense of obligation to the organization” (Allen & Meyer, 1996,

p. 253).

Organizational commitment: Organizational commitment refers to a multi-

dimensional construct consisting of “a psychological link between the employee and his

or her organization that makes it less likely that the employee will voluntarily leave the

organization” (Allen & Meyer, 1996, p. 252).

Psychological capital: Psychological capital is a positive psychological state of

development where hope, self-efficacy, resilience, and optimism work synergistically to

create internal motivations, perseverance, and positive emotions (Luthans & Youssef-

Morgan, 2017).

Resource caravans: This is a theory that personal resources, including

psychological resources, “do not exist individually but travel in packs, or caravans, for

both individuals and organizations (Hobfoll et al., 2018).

Resource caravan passageways: Included within Hobfoll’s COR theory, resource

caravan passageways are a set of environmental conditions which may encourage or

impede the development of personal resources and resource caravans (Hobfoll et al.,

2018).

Assumptions

In the development and execution of this study, I made the following

assumptions. First, I assumed that using an online survey tool was the most efficient and

effective method of collecting data for this study. Additionally, I assumed that employees

would be familiar with completing an online survey tool as the organization has used

similar tools in the past for their data collections. Second, I assumed that participants

were truthful and honest in their responses, and they were not engaging in any intentional

efforts to mislead or distort the study’s findings. Third, I assumed that the instruments

used to collect data are reliable and valid and accurately measure cross-cultural PsyCap

and OC.

Scope and Delimitations

Scope of the Study

The scope of this study included the employees at a large health care organization

in Alberta, Canada. The research population included all staff, including

corporate/business staff (i.e., human resources, finance, strategic planning), maintenance

and support services (i.e., facilities management, environmental and nutrition services),

and front-line staff (i.e., registered nurses, care support, unit clerks). The research sample

consisted of those employees who responded to the survey. The minimum sample size

suggested by Soper (2020) is 138; however, Memon et al. (2020) recommend that for

proper data analysis, the sample should be between 160 and 300. The data collected from

the sample were limited to demographic data, the responses to the cross-cultural PsyCap

scale, and the three-component model employee commitment survey. The cross-cultural

PsyCap scale was previously used in research by Reichard et al. (2014), Kotze and

Massyn (2019), and Maslacki and Sesen (2019). Numerous studies use the three-

component employee commitment survey (Allen & Meyer, 1996; Meyer & Allen, 1991;

Meyer & Herscovitch, 2001; Meyer et al., 2002).

Delimitations of the Study

There are several delimitations or boundaries within this study. First, although I

sent the digital invitation to all staff at the organization, there was no way to verify if all

staff received the invitation or which employees were willing to participate without

risking the confidentiality and anonymity of the participant responses. Additionally,

while this study is primarily concerned with the relationship between cross-cultural

PsyCap and OC, cross-cultural PsyCap is one factor out of many that may influence OC.

Furthermore, the literature states that cross-cultural PsyCap is a higher-order construct

consisting of cross-cultural hope, efficacy, resilience, and optimism. However, Luthans

and Youssef-Morgan (2017) have suggested this is not a finite list, and there may be

other psychological resources that meet the characteristics required under PsyCap. These

other psychological resources, such as mindfulness, gratitude, and courage, are outside

this study’s boundaries.

Limitations

There are certain limitations present in this study. First, there is a limitation

concerning the generalizability of the study findings. The study used participants from

one health care organization within a single province in Canada. While results may be

generalizable within the province, further studies would be required to extend the

generalizability of the results beyond those boundaries. Additionally, the health care

organization would be considered a large, public-sector organization; as noted by

Maslakci and Sesen (2019), intercultural interactions may differ in small, medium, or

private-sector organizations. A second limitation is that the study used self-reported

measures and, as such, was exposed to the potential of social desirability bias. As the

survey asked participants to provide beliefs and attitudes about situations and interactions

involving culturally diverse groups of individuals, there may be potential for responses to

reflect how the individual believes the organization would want them to respond (i.e.,

what is socially acceptable). I made all possible efforts to ensure participants’ anonymity

to allow for open and honest participation to address this limitation. A third limitation

may include the use of online surveys for data collection. While measures to preserve

anonymity were a key factor, such measures may prevent verifying that participants

understand the survey questions and responses. The online survey interface may also be a

barrier for those unfamiliar with responding to survey questions in a digital environment.

Significance

This research fills a gap in understanding cross-cultural PsyCap and how cross-

cultural PsyCap influences OC in multicultural organizations. Kotze and Massyn (2019)

indicated that interactions between different cultures within the workplace could be

emotionally draining and deplete employees’ psychological resources; however, cross-

cultural PsyCap comprises positive psychological resources that can reduce or eliminate

many of the negative experiences. Increased cross-cultural PsyCap may make working in

a culturally diverse organization a more positive experience for all employees.

Additionally, understanding the nature of the relationship between cross-cultural PsyCap

and OC may provide additional insight into achieving positive organizational outcomes

(i.e., OC) through supporting employees in developing cross-cultural skills and

competencies.

The results of this study will also provide insight into how positive organizational

outcomes can be achieved through training and developing employees. Multiple studies

have shown that employees can build cross-cultural PsyCap through focused training and

development sessions (Dollwet & Reichard, 2014; Kotze & Massyn, 2019). Such insights

will support the notion that organizations can develop competitive advantages by

supporting and developing personal resources in employees.

The present study provides an original contribution to the current literature by

focusing specifically on the nature of the relationship between the personal psychological

resources that work together to form cross-cultural PsyCap and the affective, normative,

and continuance components of OC in a multi-cultural employment setting. This research

also contributes to the industrial and organizational psychology literature and, more

specifically, to PP and POB.

Summary and Transition

In Chapter 1, the constructs of PsyCap, cross-cultural PsyCap, and OC were

summarized, particularly in the context of being psychological states which are valuable

and desirable to both employees and organizations. This study also highlighted the

importance of the relationship between cross-cultural PsyCap and positive work

outcomes within a multicultural environment. As globalization has changed the

demographics of workforces worldwide, successful intercultural interactions have

become crucial to individual performance and a vital component of establishing a

competitive advantage in a global context. Chapter 1 also emphasized the importance of

understanding how domain-specific and state-like constructs operate within specified

contexts and stated this study’s contributions to the current literature. Four specific and

appropriate research questions were identified, aligning the research’s background,

problem statement, nature, and purpose. The chapter introduced COR theory to help

convey how the constructs function and interact with each other. Finally, the chapter

presented essential definitions and assumptions and provided several limitations which

underpin the research.

Chapter 2 summarizes the research and literature related to PsyCap, cross-cultural

PsyCap, COR theory, and OC. The chapter provides an introduction that summarizes the

constructs and theories in this study, followed by a summary of the literature search

strategy used for the chapter. COR theory is discussed and presented as a theoretical

framework to understand the variables of this study. The chapter then transitions to a

discussion of PsyCap, including the historical development and the relationship between

PsyCap, POB, POS, and PP. The chapter then moves into an overview on cross-cultural

PsyCap, discussing the adaptation of the PsyCap measurement tools, results of research

specific to cross-cultural PsyCap, and eventually summaries of each of the individual

psychological resources which form the construct. Finally, an overview of OC is

provided, including a discussion on the development of the three-component model of

OC, including continued model modifications and critiques, which some researchers have

provided. The chapter offers individual summaries on each component of OC and closes

by summarizing the significant constructs and restating the literature gap, which this

study fills.

Chapter 3 provides a detailed discussion on the methodology used within this

study. The study design is discussed in detail within the chapter and explains the

population and sampling procedures. The chapter provides the rationale for using

structured equation modeling (SEM) and explains how the data are collected and how it

will be analyzed to test the hypotheses. Finally, the chapter presents the tools used to

measure the constructs and operationalization of essential concepts before summarizing

the threats to validity.

Chapter 2: Literature Review

Introduction

The higher-order concept of cross-cultural PsyCap has been adapted from earlier

research on PsyCap by Luthans et al. (2007a). However, cross-cultural PsyCap exists as a

context-specific form of PsyCap focused on intercultural interactions within an

organization (Dollwet & Reichard, 2014; Kotze & Massyn, 2019). While researchers to

date have contributed significantly to the concept of various context-specific forms of

PsyCap, they have only scratched the surface regarding illuminating how organizations

can leverage cross-cultural PsyCap to create organizational competitive advantage (Kotze

& Massyn, 2019; Maslakci & Sesen, 2019). Specifically, researchers describe cross-

cultural PsyCap as a “tool for the development, measurement, and effective management

of the positive behaviors of employees in organizations” (Maslakci & Sesen, 2019, pg.

65); which in and of itself speaks to the potential the concept may hold for organizations

to leverage for competitive advantage.

Whereas the literature identifies cross-cultural PsyCap as a higher-order construct,

the construct has a more substantial cumulative effect than any individual components

upon which cross-cultural PsyCap is built (Luthans et al., 2007a). Dollwet and Reichard

(2014) incorporated this concept of a cumulative effect by adapting the earlier identified

variables for workplace PsyCap to represent the intercultural interactions that occur

within multicultural organizations. In Dollwet and Reichard’s adaptation, cross-cultural

hope, cross-cultural self-efficacy, cross-cultural resilience, and cross-cultural optimism

contribute to overall cross-cultural PsyCap. The individual components reflect an

employee’s ability to effectively function within a multicultural workplace and with

colleagues who represent diverse and non-diverse cultures. In this conceptualization,

cross-cultural hope generally describes the ability of employees to set and achieve goals

in a multicultural environment. Cross-cultural self-efficacy describes significant

confidence in one’s ability to communicate and adapt in a multicultural setting. Cross-

cultural resilience describes an ability to rise above differences in language or other

cross-cultural difficulties and continue to perform. Finally, cross-cultural optimism

defines the ability to remain positive about current and future intercultural interactions

(Kotze & Massyn, 2019).

OC has been well-represented in the current literature; however, while many

conceptualizations of OC exist, one of the most popular and enduring models has been

Meyer and Allen’s (1991) three-component model of OC. The model proposed by Meyer

and Allen (1991) consists of three dimensions, namely AC, NC, and CC. The model

intends to cover both the attitudinal and behavioral approaches to OC (Meyer & Allen,

1991), outlining both the attachment an employee has developed to an organization and

their willingness to remain employed with the same employer (Sen et al., 2017). To

outline this point, AC describes an emotional attachment to an organization, where

commitment relates to a perception of shared values. NC represents a moral attachment,

where commitment reflects loyalty. Finally, CC describes a continuation of a current

attachment, where commitment includes an evaluation of suitable alternatives (Meyer &

Allen, 1991; Sen et al., 2017). Previous literature has identified a significant relationship

between OC and turnover intent (Sen et al., 2017; Seo & Chung, 2019). However,

organizations that maximize OC while minimizing turnover intent benefit from a more

engaged workforce and reduced costs associated with recruit and training, which can set

the groundwork for organizational competitive advantage (Basit, 2018; Peng et al., 2013).

Literature Search Strategy

I started my literature search using the Thoreau Multi-Database Search available

through the Walden University Online Library. The initial search terms I used

were psychological capital or PsyCap, which returned 5,632 results. I refined the results

by filtering to include only peer-reviewed scholarly journal articles, only academic

journal publication types and using the date range filter to limit results to those published

between 2015 and 2021, which returned 3,388 results. For the second search attempt, I

used the search terms cross-cultural psychological capital or cross-cultural PsyCap. I

repeated the same date range and limiters; this search returned five results. I also used the

search terms psychological capital, and cross-cultural psychological capital with

the PsycArticles, PsycBooks, PsycExtra, PsycInfo, PsycTests, and Business Source

Complete databases; however, the data range was expanded from 2014 to 2021, returning

672 and seven results, respectively. Both psychological capital and cross-cultural

psychological capital search terms were next used on Google Scholar, using a date range

of 2015 to 2021. Both search terms returned a significant number of results (3,260 and

670, respectively).

While the initial literature searches used single search terms, subsequent searches

used the cross-cultural psychological capital search term combined with OC, cultural

competence, hope, self-efficacy, resilience, and optimism, yielding mixed results. I

subsequently conducted multiple targeted searches based on relevant research articles and

theoretical frameworks identified in the research studies obtained from the initial search

efforts. Inclusion criteria for literature in this review were as follows: content explicitly

related to cross-cultural PsyCap, content about PsyCap, content specific to relationships

with positive organizational/work outcomes, and content specific to employees in work

settings.

Conservation of Resources Theory

Hobfoll’s (1989) COR theory is a stress and motivation theory that Hobfoll

originally conceptualized as a testable model which explains the “ubiquitous stress

phenomena and perhaps bridges the gap between environmental and cognitive

viewpoints” (p. 516). The fundamental assertion in COR theory is that individuals will

use essential resources to pursue, maintain, and protect those things, including resources,

which they genuinely value (Hobfoll, 2011). Hobfoll (2011) suggested a universal nature

exists within COR theory to provide more clarity regarding the concept of things of

value. Hobfoll (2011) indicated that which individuals genuinely value “includes health,

well-being, peace, family, self-preservation, and a positive sense of self, even if the core

elements of sense of self differ culturally” (p. 117).

Beyond the fundamental aspects, four main principles underpin COR theory

(Hobfoll, 1989; 2011). The first principle argues that resource loss is significantly more

impactful than resource gain; for example, the “primacy of resource loss” (Hobfoll, 2011,

p. 117) principle suggests that from a psychological perspective, a reduction in

compensation is more impactful than an equivalent increase in compensation

(Halbesleben et al., 2014). The primacy of resource loss principle becomes even more

relevant when a specific resource is finite. Halbesleben et al. (2014) suggested in such

scenarios, the loss of resources could be devastating as the “resource gain cycle”

(Hobfoll, 2011) is difficult to initiate due to the limited resources. The second principle

echoes the impact of the primacy of resource loss principle as the second principle asserts

that one must invest resources to pursue new resources, to protect against loss, and should

loss occur, to recover from such loss (Hobfoll, 2011). If the individual has invested

significant resources to protect against such loss, it may significantly exasperate the

impact of initial resource loss. Alternatively, the individual may also invest resources to

gain additional resources. However, the resultant gain of resources may buffer the

resources lost through investment. The third principle, the gain paradox principle

(Hobfoll et al., 2018), suggests that when resource losses already exist, resource gain

becomes significantly more important. Initially, the gain paradox principle may appear to

conflict with the first two principles of COR theory. However, resource loss would

remain more psychologically impactful than resource gain. Once losses have occurred, an

individual is more likely to deploy resources that focus on resource gain to recover from

the loss (i.e., resource investment), creating a somewhat paradoxical scenario. Finally, the

fourth principle describes a state of desperation. When resources are at significant risk of

exhaustion, individuals may display irrational, aggressive behavior in defense of their

resources (Hobfoll et al., 2018).

Three corollaries are also applicable within COR theory (Halbesleben et al., 2014;

Hobfoll, 2011; Hobfoll et al., 2018). The first corollary identifies that individuals with

greater resources can better invest resources to buffer themselves against loss or propel

them towards resource gain. Alternatively, individuals with fewer resources have less

ability to deploy resources to prevent or protect against loss, allowing further losses to

occur. The second corollary suggests a spiraling effect to the resource loss cycle (Hobfoll

et al., 208). As described in the first principle of COR theory, resource loss is

psychologically more impactful than resource gain; Hobfoll (1989; 2011) suggested that

the disproportionate effect of resource loss is due to the additional stress. As such, with

each successive loss of resources (i.e., loss spiral), fewer resources are available to

protect against further loss, and thus the spiral gains momentum (Hobfoll et al., 2018).

The third corollary suggests that just as resource loss has a spiraling effect, there is a

spiraling effect to resource gain (Hobfoll et al., 2018). Unfortunately, whereas resource

gain is somewhat more challenging to achieve and significantly less impactful (i.e.,

principle 1), resource gain spirals tend to be more difficult, sometimes monotonous and

can be exceptionally slow processes (Hobfoll, 2011; Hobfoll et al., 2018).

Modern conceptualizations of COR theory also propose the existence of resource

caravans and resource caravan passageways. The caravans and passageways provide “a

greater understanding and emphasis on both the interrelationship between resources and

how environments and contexts create fertile or infertile ground for creation,

maintenance, and limitation of resources” (Hobfoll et al., 2018, p. 107). COR theory

proposes that resources do not exist separately and independently of other resources;

instead, resources tend to travel in packs or resource caravans with additional resources

and synergistic effects (Halbesleben et al., 2014; Hobfoll, 2011; Hobfoll et al., 2018).

The conceptualization of the resource caravan passageway is, at the core, a recognition

that individuals, and their resources, are influenced by the societal and ecological

conditions in which they exist (Halbesleben et al., 2014; Hobfoll, 2011; Hobfoll et al.,

2018). Hobfoll (2011) proposed that such environmental conditions may “support, foster,

enrich, and protect the resources of individuals, sections or segments of workers, and

organizations in total, or that detract, undermine, obstruct, or impoverish people’s or

group’s resource reservoirs” (p. 118-119). This level of analysis may become acutely

important within workplace settings as the organization and the organizational culture

play significant roles in creating the societal and ecological conditions in which

employees function.

Whereas COR theory is a stress and motivation theory, several types of resources

are discussed based on the likelihood that resource gain or resource loss results in stress

or well-being. Object resources are physical resources such as tools or a car (Hobfoll,

2011), which require resource investment to gain (i.e., purchase) and also require

resource investment to prevent against loss (i.e., dilapidation). The personal value that an

individual assigns to object resources would be closely related to the acquisition or

replacement cost of the item (Hobfoll, 1989). Conditions or condition resources are terms

that have been used in COR theory to describe external situations or social scenarios

which are considered desirable for individuals due to “stress-resistance potential”

(Hobfoll, 1989, p. 517). Condition resources could include marriage and tenure (Hobfoll,

1989) or even seniority and positive relationships within the workplace (Hobfoll, 2011).

Energies, or energy resources, are often valued as resources due to the energies required

in acquiring, maintaining, and protecting other resources (Hobfoll, 1989). It is relatively

straightforward that there must be an investment of time, effort, or money (i.e., energy

resources) in the collection of or protection of object, condition, or personal resources.

However, as Hobfoll (1989) suggested, the value of energy resources closely relates to

the stress derived from their loss, often through investment. For example, wasted time or

effort that did not result in gain, maintenance, or protection of other resources would be

felt disproportionately as a loss, as suggested by the first principle of COR theory. The

final resources in COR theory, and arguably the most relevant to PsyCap and cross-

cultural PsyCap, are personal characteristics (Hobfoll, 1989), sometimes referred to as

personal resources, skills, or traits (Hobfoll, 2011). Personal resources in this light would

include the personal psychological resources (i.e., hope, efficacy, resilience, and

optimism) which form and travel together within the PsyCap construct (Kotze & Massyn,

2019; Liu, 2014; Mao et al., 2020). Hope, efficacy, resilience, and optimism travel

together (i.e., resource caravans) and show synergistic effects (Luthans & Youssef-

Morgan, 2017). Through COR theory, we understand that such resources are valuable to

individuals who would then invest further resources in pursuing, maintaining, and

protecting these resources (Hobfoll, 1989; 2011). The principles and corollaries in COR

theory also help explain how and why individuals deploy personal psychological

resources (i.e., cross-cultural PsyCap) when presented with potentially stress-inducing

interactions with people from different cultures in a multicultural workplace setting

(Kotze & Massyn, 2019). Consistent with COR theory, Sungu et al. (2020) suggested that

individuals will deploy significant resources when the deployment is likely to gain or

protect resources they genuinely value. Therefore, if OC is considered a desirable

condition involving emotional attachment to the organization, COR theory would suggest

that employees need resources they can deploy to support OC. These resources which

support OC may include personal psychological resources such as those embedded in

cross-cultural PsyCap. However, Wright and Hobfoll (2004) indicated that two other

conditions must be present. In the first condition, which Sungu et al. (2020) also refer to,

the individual must possess the required resources for deployment. In the second

condition, the environment or organization must present the opportunity for individuals to

deploy their resources. For example, an individual with high cross-cultural PsyCap (i.e.,

available personal resources) would need the opportunity to deploy such resources in

situations requiring interacting with people from different cultures to support the gain,

maintenance, or prevention of loss of OC.

Psychological Capital

Initially conceptualized as a specific construct to be included under POB, PsyCap

can trace its roots to the broader movements of POS and PP (Luthans & Youssef-

Morgan, 2017). PP is a field of study founded in the late nineties, which intends to focus

on underrepresented areas in the annals of clinical psychology. More specifically, the

focus was on the positive aspects of life, including “positive experience, positive

institutions, and positive traits” (Seligman, 2019). While PP intended to divert research

towards a field that considers the positive aspects of individuals (Seligman, 2019), there

was also a need within such a field to apply this research focus in the domain of

organizational science, which became known as POS. Cameron (2017) described POS as

“an umbrella framework used to unify a variety of concepts in organizational studies,

each which incorporates the notion of the positive” (p. 13). In terms of what is considered

the positive within the cross-cultural literature, Cameron further identified that

researchers of POS generally expect several characteristics of anything included within

the framework of POS. Expectations include positive deviance in which an outcome of

the deviance dramatically exceeds the usual or expected outcome. The characteristics also

include affirmative bias, where the intent is to focus on positive traits rather than negative

factors, and virtuousness or eudemonism, where the positive has intrinsic value and is not

just a pathway to achieve something else. Whereas PP and POS have a wide range of

applications, POB is concerned with the characteristics of specific constructs. Luthans

and Youssef-Morgan (2017) suggested that a construct must meet a broad set of attributes

to align with PP and POS. However, there is a more specific set of characteristics that

constructs must meet to align with POB. These characteristics include that theory and

evidence provide the basis to the construct. The orientation of the construct is a

positioning that is positive within the domains of PP and POS. The construct must be

scientifically measurable. The construct must present the opportunity for further

development, and finally, the construct must relate to desirable work outcomes (Luthans

& Youssef-Morgan, 2017).

Aligning with PP, POS, and POB, PsyCap is:

an individual’s positive psychological state of development that is characterized

by: (1) having confidence (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

necessary, 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., 2015, p. 2)

As a higher or second-order construct, a combination of personal psychological

resources, including hope, efficacy/self-efficacy, resilience, and optimism, form PsyCap

and are often referred to by the acronym HERO. Further aligning with the requirements

to be included under POB, PsyCap exists as a state-like resource. The state-like

conceptualization means that PsyCap is more constant than certain emotional states but

less stable than pure traits. There is a flexibility to the construct, which allows the

construct to be further changed and developed through experiences and education

(Luthans & Youssef-Morgan, 2017). Additionally, PsyCap is a context-specific or

domain-specific construct (Dollwet & Reichard, 2014; Luthans & Youssef-Morgan,

2017). Individuals may have high levels of PsyCap (i.e., hope, self-efficacy, resilience,

and optimism) related to their specific workplace, job, or tasks. However, the same

individuals may have lower levels of PsyCap when it comes to other contexts or domains.

These domains may include relationships and health (Luthans et al., 2013) or intercultural

interactions (Dollwet & Reichard, 2014).

Cross-Cultural Psychological Capital

To address the context-specificity of PsyCap and extend the construct into the

domain of intercultural interactions, Dollwet and Reichard (2014) proposed the concept

of cross-cultural PsyCap. Rather than focusing specifically on workplace PsyCap, cross-

cultural PsyCap instead was focused on psychological resources which span cultural

differences and contribute to successful intercultural interactions, including those which

happen within the workplace (Dollwet & Reichard, 2014). This definition of cross-

cultural PsyCap differs from Yunlu and Clapp-Smith’s (2014) conceptualization of

cultural PsyCap, which extends the PsyCap construct into the domain of cross-cultural

experiences related to relocating and working in foreign countries. While both Dollwet

and Reichard (2014) and Yunlu and Clapp-Smith (2014) started with Luthans et al.’s

(2007a) Psychological Capital Questionnaire (PCQ), they independently developed

separate scales for cross-cultural PsyCap and cultural PsyCap, respectively. Whereas

Luthans et al.’s (2007a) original PCQ used terminology such as ‘I always look on the

bright side of things regarding my job,’ Dollwet and Reichard (2014) used the wording ‘I

always look on the bright side of things regarding my cross-cultural interactions.’ Yunlu

and Clapp-Smith (2014) used the terminology ‘I always look on the bright side of things

regarding what I experience in other cultures,’ highlighting that while conceptually

similar, the measures are focused on different constructs.

In conceptualizing cross-cultural PsyCap, Dollwet and Reichard (2014) initially

conducted two studies to validate the construct. The initial research uses an online survey

to collect data from a diverse group of participants (N = 361) recruited through the

Mechanical Turk (Mturk) database. In the second study, Dollwet and Reichard collected

data from another 134 participants. The researchers employed a sampling strategy similar

to the initial research to ensure a demographically diverse participant group. The initial

investigation confirmed the higher-order conceptualization of cross-cultural PsyCap. The

second study analyzed the reliability and validity of the construct in terms of specific

positive organizational outcomes, including cultural intelligence, cross-cultural

adjustment, ethnocentrism, and openness to experience. After removing nine items from

the initial cross-cultural PCQ for poor fit, Dollwet and Reichard reported high

comparative and incremental fit indices (CFI = 0.91, IFI = 0.91), root mean square error

of approximation (RMSEA) = 0.08, and Cronbach’s alpha of 0.79 to 0.91 across all sub-

scales, indicating both the goodness of fit and the superiority of the four-factor model of

cross-cultural PsyCap. Similar fit indices were reported in study 2 (CFI = 0.88, IFI, 0.88),

indicated a continued fit between the model and the data (Dollwet & Reichard, 2014). In

addition to assessing fit, Dollwet and Reichard conducted regression analyses that

indicated statistically significant positive relationships between cross-cultural PsyCap,

cultural intelligence, openness to experience, and cross-cultural adjustment and a

statistically significant negative relationship with ethnocentrism.

Building on the previous cross-cultural PsyCap research, Reichard et al. (2014)

also focused on the relationship between cross-cultural PsyCap, cultural intelligence, and

ethnocentrism; however, this study introduced a training intervention. The researchers

obtained data from 130 participants from various organizations in California and another

71 participants employed by a university in South Africa (Reichard et al., 2014). The

researchers used a pretest-posttest design in which a pre-intervention survey was initially

completed. The researchers then provided participants with a two-hour training session

“focused on creating self-awareness, reframing past events, building broad cross-cultural

interaction skills, and identifying multiple strategies for success in cross-cultural

interactions” (Reichard et al., 2014, p. 155). Finally, participants completed post-

intervention surveys. Similar to the initial studies on cross-cultural PsyCap, Reichard et

al. assessed for model fit; fit indices for the US group (CFI = 0.97, IFI = 0.97) and the

South African group (CFI = 0.98, IFI = 0.98) were high. All factor loadings for hope,

self-efficacy, resilience, and optimism were significant, indicating a good fit. The

researchers also performed paired-samples t test and repeated measures ANOVA,

indicating statistically significant increases in cross-cultural PsyCap and cultural

intelligence and decreases in ethnocentrism from pretest to posttest. The research by

Reichard et al. was novel due to the inclusion of a training intervention, which indicated

that organizations could increase employees’ cross-cultural PsyCap and related positive

organizational outcomes through relatively short bursts of training. Additionally, the

research supported previous PsyCap intervention research (Luthans et al., 2010; Luthans

et al., 2014) and extended it into the domain of intercultural interactions.

Subsequent research using Dollwet and Reichard’s (2014) cross-cultural PsyCap

measure has shown that positive relationships exist between the construct and specific

positive organizational outcomes. In a study using 213 employees from various South

African organizations, Kotze and Massyn (2019) found a statistically significant positive

relationship between cross-cultural PsyCap and work engagement and a statistically

significant negative relationship with burnout. Using the vigor and dedication

components of work engagement and the cynicism and emotional exhaustion components

of burnout, Kotze and Massyn used partial least squares (PLS) and SEM not only to test

the relationships with cross-cultural PsyCap but also to test for internal

consistency/composite reliability and construct validity (i.e., convergent and discriminant

validity). Composite reliability scores for each construct in the model ranged from 0.842

to 0.934. The outer loadings ranged from 0.656 to 0.946, and the average variance

extracted ranged from 0.577 to 0.826 indicating acceptable convergent validity (Kotze &

Massyn, 2019). To test for discriminant validity, Kotze and Massyn used the heterotrait-

monotrait (HTMT) ratio of correlations as suggested by Henseler et al. (2015). After

removing two items due to high correlations between vigor and dedication and one due to

a high correlation between cross-cultural self-efficacy and cross-cultural resilience, the

researchers achieved the desired HTMT ratio of less than 0.85. Kotze and Massyn’s

results were significant as the findings provided context-specific support to previous

studies involving the relationships between PsyCap, work engagement, and turnover (Du

Plessis & Boshoff, 2018; Kotze, 2018a, 2018b; Peng et al., 2013).

Maslakci and Sesen (2019), similar to Kotze and Massyn (2019), also explored

the relationship between cross-cultural PsyCap and positive organizational outcomes. In a

study using 346 employees from several different five-star hotels in Northern Cyprus,

Maslakci and Sesen identified that cross-cultural PsyCap positively mediates the

relationship between multicultural personality traits and perceived service quality. The

researchers used confirmatory factor analysis (CFA) to test the scales’ validity and used

SEM for hypothesis testing (Maslakci & Sesen, 2019). As suggested by Schreiber et al.

(2006), the researchers reported common fit indices, goodness of fit (GFI) = 0.91, CFI =

0.90, IFI = 0.89, and RMSEA = 0.06 for the model. Maslacki and Sesen’s findings

aligned with previous research indicating the positive nature of PsyCap in the service

industry (Bouzari & Karatepe, 2017; Kim et al., 2018). However, the findings also

supported research indicating that cross-cultural PsyCap had similar positive

relationships with desirable organizational outcomes as workplace PsyCap (Dollwet &

Reichard, 2014; Kotze & Massyn, 2019; Reichard et al., 2014).

The scales that Dollwet and Reichard (2014) adapted from Luthans et al.’s

(2007a) earlier work utilized the same personal psychological resources present in the

PCQ. However, Dollwet and Reichard conceptualized the psychological resources as

domain-specific resources and extended the concept into the domain of intercultural

interactions. As such, the researchers conceptualized HERO as cross-cultural HERO,

including cross-cultural hope, cross-cultural efficacy/self-efficacy, cross-cultural

resilience, and cross-cultural optimism (Dollwet & Reichard, 2014; Reichard et al.,

2014).

Cross-Cultural Hope

Hope initially defined a personal psychological resource related to a motivational

state consisting of agency, pathways, and goals (Luthans et al., 2007a). However, most

research (Dollwet & Reichard, 2014; Khandelwal & Khanum, 2017; Newman et al.,

2014; Reichard et al., 2014) has focused mainly on agency and pathways, as agency and

pathways constructs subsume the goal construct. Much of our understanding of agency

and pathways relies on the work of Snyder et al. (1996), whose research validated a

measure of hope as a state consisting of the agency and pathway components. In this

context, agentic thinking was the motivational aspect (Snyder, 2002), the determination

to reach a goal (Snyder et al., 1996), and confidence in using one’s pathways to achieve it

(Snyder, 2002). While pathways are primarily the means through which an individual

achieves a goal (Snyder et al., 1996), pathway thinking refers to considering alternative

options or pathways and the confidence in selecting a route (Snyder, 2002). Summarizing

the difference between the agency and pathways concepts and agentic and pathway

thinking, Khandelwal and Khanum (2017) suggested, “Thus, hope is not just the positive

anticipation but also having plans to achieve the goals.” (p. 89). This differentiation

clarifies that hope is not just general positivity but rather the embedding of positivity in

the actions taken towards achieving goals.

Hope, as a first-order component of PsyCap, has been positively related to many

positive work outcomes, including work performance (Reichard et al., 2013), job

satisfaction (Badran & Youssef-Morgan, 2015; Jung & Yoon, 2015; Olaniyan & Hystad,

2016; Sen et al., 2017; Tang et al., 2019), OC (Hussain & Nawaz, 2019; Sen et al., 2017;

Tang et al., 2019; Yildiz, 2018), job involvement (Demir, 2018), psychological well-

being (Avey et al., 2010; Reichard et al., 2013), organizational citizenship (Bogler &

Somech, 2019; Jung & Yoon, 2015), and negatively related with stress (Demir, 2018;

Hussain & Nawaz, 2019; Sen et al., 2017), anxiety (Demir, 2018; Zhou et al., 2018), and

burnout (Demir, 2018; Kotze, 2018a; Peng et al., 2013; Zhou et al., 2018). However,

inconsistent with the results of the other researchers, Idris and Manganaro (2017) were

not able to find a relationship between PsyCap and either job satisfaction or OC; they

hypothesized the contradictory results may have reflected a difference in cultural

practices and language barriers.

Whereas there is a significant volume of research including hope as a component

of PsyCap, there is significantly less research extending the psychological construct of

hope into the domain of cross-cultural interactions. Building on the earlier studies by

Snyder et al. (1996), Luthans et al. (2007a), and others, Dollwet and Reichard (2014)

conceptualized cross-cultural hope as “pursuing and meeting goals related to working

with people from different cultures” (p. 1672). As such, individuals with high cross-

cultural hope are more likely to be positively motivated to interact with individuals from

different cultures. These individuals may also be better able to identify and select

pathways to avoid or overcome any problems which may impede their ability to interact

with other individuals (Dollwet & Reichard, 2014; Reichard et al., 2014). The ability to

“produce plausible alternative routes” (Snyder, 2002, p. 251) underpins the assertion that

high cross-cultural PsyCap better equips individuals to interact cross-culturally.

Particularly when language barriers, differences in cultural norms, and lack of

information are routine (Dollwet & Reichard, 2014; Reichard et al., 2014), the individual

can remain positive about the interaction and devise other pathways to reach success.

Cross-cultural hope is essential in culturally diverse organizations. Whereas the positive

organizational outcomes related to hope have been well-documented (Newman et al.,

2014), researchers have postulated that cross-cultural hope would be associated with

similar outcomes (Reichard et al., 2014). However, cross-cultural hope may also create

an awareness of cultural assumptions and biases (Reichard et al., 2014) within the

organization; addressing such assumptions and biases would be vital to leverage cross-

cultural hope, and more broadly, cross-cultural PsyCap for organizational success.

Cross-Cultural Efficacy/Self-Efficacy

Often used interchangeably, efficacy or self-efficacy refers to individuals’

confidence in their abilities to take on new challenges and mobilize their efforts towards

completing those challenges successfully (Luthans & Youssef-Morgan, 2017). Based on

the work of Bandura (1997), these personal expectations will determine the motivation

and psychological resources that an individual invests into achieving their goal and how

likely they are to quit should they run into barriers (Stajkovic & Luthans, 1998). Similar

to hope, Bandura suggested that self-efficacy is rooted in human agency. The decisions

made and actions taken are intentional; the individual is the ‘agent’ of their success or

failure. Whereas Stajkovic and Luthans (1998) took the concept of self-efficacy and

redefined it in terms of employees and a specific task within the context of the workplace,

Luthans et al. (2007a) further refined the definition and extended the concept beyond the

completion of individual tasks to a broader work domain in the conceptualization of

PsyCap. For example, within the work context, an individual with high self-efficacy

would be expected to have a higher level of confidence in their ability to complete a set

or group of related tasks explicitly related to their position or role (Luthans et al., 2007b;

Newman et al., 2014).

There is a positive link between self-efficacy and work-related performance, both

when self-efficacy is an independent construct (Stajkovic & Luthans, 1998) and as a core

component of PsyCap. Aligning with Hobfoll’s (2011) COR theory, researchers

investigating self-efficacy as a component of PsyCap have found often there are

synergies when the psychological resources (i.e., hope, self-efficacy, resilience, and

optimism) travel together (Luthans & Youssef-Morgan, 2017). Stronger correlations are

found between the higher-order PsyCap construct and various positive organizational

outcomes when compared to the correlations between individual psychological resources

(i.e., self-efficacy) and the same desirable outcomes (Luthans et al., 2007a). In other

cross-cultural research not directly related to PsyCap, Luthans, Zhu, and Avolio (2006)

found that self-efficacy was positively associated with OC and negatively associated with

turnover intent. Furthermore, the findings also supported their hypothesis that the

relationship would be more robust in individualistic cultures when compared with

collectivist cultures.

Whereas self-efficacy describes individuals’ belief in their abilities (Luthans &

Youssef-Morgan, 2017), cross-cultural self-efficacy describes individuals’ belief in their

ability to interact with others. Including interaction with co-workers from different

cultures (Dollwet & Reichard, 2014) and their confidence in their ability to employ

various methods to make those interactions successful (Nunez, 2000). In this context,

cross-cultural self-efficacy is more than simply knowing and understanding different

cultures (Dollwet & Reichard, 2014). It focuses more on mutually successful interactions

in diverse settings rather than merely delineating how one culture is different (Nunez,

2000). As such, individuals with high cross-cultural self-efficacy would feel more

comfortable and confident in multicultural work settings (Dollwet & Reichard, 2014).

They would feel confident in their ability to adapt to the needs of each situation (i.e.,

intercultural interaction) and motivated to find new ways of continuing to adapt (Nunez,

2000; Rehg et al., 2012). However, it is important to note that cross-cultural self-efficacy

is related to individuals’ beliefs about their abilities (Dollwet & Reichard, 2014; Luthans

&Youssef-Morgan, 2017) rather than their actual ability. Individuals may become

overconfident in their ability to interact with others (e.g., belief that all interactions with

people of a particular culture will be the same). Negative self-perception and previous

failures may also distort individuals’ evaluation of their abilities and result in feelings of

incompetence (Bandura, 1997; Stajkovic & Luthans, 1998). As these misjudgments about

perceived versus actual ability are less likely to occur when the individuals are likely to

experience significant consequences due to their over or under-confidence (Bandura,

1997), organizations can help individuals close any gaps (Reichard et al., 2014). As a

component of cross-culture PsyCap, cross-cultural self-efficacy is both state-like and

malleable (Luthans et al., 2006; Luthans & Youssef-Morgan, 2017; Reichard et al.,

2014), and individuals can increase cross-cultural self-efficacy through several methods.

Bandura (1997) outlined several methods, including developing mastery (i.e., actual

ability) through successful experiences, vicarious learning through watching others model

success, social persuasion through positive encouragement and feedback from relevant

sources, and increasing positivity and psychological well-being.

Cross-Cultural Resilience

There is an extensive volume of research on the concept of resilience which

covers a variety of different domains, including developmental psychology (Luthans &

Youssef-Morgan, 2017), military and fitness training (King et al., 2016), and diverse

work environments (Meng et al., 2019). As a result, researchers developed many different

definitions for resilience in these contexts, including some that ascribe trait-like

properties to resilience (Jackson et al., 2007) and others that consider resilience to be

more of a state-like phenomenon open to development (Luthar et al., 2000). Such

discrepancies led Luthar et al. (2000) to differentiate between the trait-like properties,

which they termed as resiliency, and the state-like properties, termed resilience,

“resiliency is a personality characteristic of the individual, whereas resilience is a

dynamic developmental process” (p. 546). In this differentiation, Luthar et al. further

identified that resilience requires an individual to experience a specific event, which can

be positive or negative; resiliency, on the other hand, does not pre-suppose such an

experience.

Building on Luthar et al.’s (2000) description of resilience as a “dynamic

developmental process” (p. 546), resilience, particularly in the workplace, can be

understood simply as an individual’s ability to adapt and respond to a specific event or

set of circumstances (Masten, 2001) within the workplace. The events or circumstances

could be significant adversities (i.e., failures); or could represent generally positive but

significant experiences (i.e., promotions), in which the individual would have to adapt to

be successful (Dollwet & Reichard, 2014; Masten, 2001). In this context, resilience is the

cumulative result of the deployment of personal assets when risk factors are present

(Dollwet & Reichard, 2012; Reichard et al., 2014). For example, individuals who have

developed high levels of resilience would strive to overcome adversity, both common and

novel types. They would effectively deploy the necessary resources or personal assets to

be successful, learning and developing higher levels of resilience in the process (Luthans

et al., 2007b).

Resilience as a developmental process, as opposed to a rare innate trait, has been

linked to many positive organizational outcomes. Avery et al. (2009) suggested the

developmental process of resilience is “arguably the most important positive resource to

navigating a turbulent and stressful workplace” (p. 682). While the concept of resilience

relates to responding positively and overcoming various physically and psychologically

stressful events (Rook et al., 2018), researchers have paid particular focus to the

relationships between resilience and stress-related burnout, turnover (Avey et al., 2009;

Jackson et al., 2007; Lee Cooke et al., 2019), and OC (Meng et al., 2019). Broadly, these

results remain consistent with earlier research on resilience which revealed a positive

influence with coping mechanisms (Masten, 2001), particularly as related to experiencing

stressful events and adversity. These characteristics of overcoming stressful events and

developing resilience towards future stressful experiences become more important in

terms of workplace resilience and the globalization of workplaces. Increased global

competition and changing workforce demographics have required organizations to

transition and embrace change which is often stressful for employees (Avey et al., 2009).

In keeping with earlier conceptualizations of resilience by Luthar et al. (2000) and

Masten (2001), cross-cultural resilience in the workplace directly relates to employees’

ability to overcome and be successful when working and interacting with people from

different cultures despite negative experiences or adversity (Dollwet & Reichard, 2014;

Reichard et al., 2014). The adaptive nature of individuals with high levels of cross-

cultural resilience (Masten, 2001) is particularly important in multicultural work settings

(Dollwet & Reichard, 2014; Reichard et al., 2014). Employees will seek to improve on

previous less successful or outright negative experiences and are more likely to deploy

the personal resources needed to maintain performance despite cross-cultural adversity

(Luthans et al., 2007a).

As a component of cross-cultural PsyCap, cross-cultural resilience is both state-

like and context-specific, meaning the development of cross-cultural resilience in the

domain of intercultural interactions may exist mutually exclusive of any other domain.

Luthans et al. (2010) suggested that individuals can develop resilience by increasing

personal assets and reducing risk factors due to the state-like nature of the construct.

Personal assets include “measurable characteristics that predict positive outcomes and

adaptation to adverse circumstances” (p.47). In contrast, risk factors include “measurable

characteristics that predict negative outcomes and poor adaptation in the workplace”

(Luthans et al., 2010, p. 47). Reichard et al. (2014) extended this idea to the domain of

intercultural interactions. They indicated individuals could develop cross-cultural

resilience by building a variety of intercultural interaction skills (i.e., personal assets) and

role-playing strategies for successful interactions (i.e., reducing risk factors).

Cross-Cultural Optimism

As a personal resource, researchers often define optimism as two independent but

related concepts. The first concept considers optimism a positive general expectancy for

any given situation (Scheier & Carver, 1992). In contrast, the second concept considers

optimism as an explanatory style (Seligman, 1998) which “attributes positive events to

personal, permanent, and pervasive causes, and interprets negative events in terms of

external, temporal, and situation-specific factors” (Luthans & Youssef-Morgan, 2017). In

dispositional optimism (Scheier & Carver, 1992), optimism and pessimism are polar

opposites, with the terms indicating either a positive (i.e., optimism) or negative (i.e.,

pessimism) general expectancy towards events (Gillhem et al., 2001). In this context,

dispositional optimism relates closely to hope and self-efficacy (Carver & Scheier, 2014),

particularly when it comes to the role of agency. Agency or agentic thinking within the

hope construct is a motivator of determination and confidence (Snyder, 2002). In the self-

efficacy construct, agency implies that individuals make intentional decisions or actions

(Bandura, 1997). Finally, in dispositional optimism, personal agency is involved in

pursuing positive expectancies and may be partially or wholly dependent on sustained

efforts (Scheier & Carver, 1992). However, as Gillham et al. (2001) argued, being

optimistic in the context of dispositional optimism could indicate more than just a general

positive expectancy for all things. Instead, being optimistic may also show an expectancy

of a certain level of control in achieving a positive outcome.

Optimism as an explanatory style (Seligman, 1998) is a concept that maintains a

significant degree of independence while still related to dispositional optimism. While

there are several differences between the two constructs, one significant divergence is the

concept of explanation rather than expectation (Gillham et al., 2001). While an individual

with high dispositional optimism may have a general expectancy of a positive outcome to

a particular event, such as a test, there is a possibility that a negative outcome (i.e.,

failure) may still occur. When presented with the same scenario, an individual with an

optimistic explanatory style would associate a positive outcome (i.e., high test score) with

personal and pervasive causes, such as their ability or aptitude. In contrast, a low score or

negative outcome would be explained by external and temporal causes, such as noise and

other distractions (Gillham et al., 2001). It is important to note; there is also a notable

difference between the explanation and expectancy concepts of optimism regarding the

role of pessimism. The current literature on dispositional optimism suggests that

optimism and pessimism hold opposite ends of a continuum (Gillham et al., 2001;

Scheier & Carver, 1992). Individuals generally expect positive events to occur or expect

adverse events to occur; the concept of dispositional optimism subsumes the concept of

pessimism (Carver & Scheier, 2014). However, researchers have begun to reconsider the

bi-polar nature of dispositional optimism. Researchers are starting to study optimism and

pessimism as independent constructs (Scheier et al., 2020), similar to how explanatory

style research views optimism and pessimism. In Seligman’s (1998) conceptualization,

optimism is a positive explanatory style where positive events are assigned internal,

stable, and global explanations and negative events are assigned external, unstable, and

specific explanations. Alternatively, a pessimistic explanatory style would assign

external, unstable, and specific explanations to positive events and internal, stable, and

global expectations to negative events (Luthans & Youssef-Morgan, 2017; Seligman,

1998). As such, while high levels of optimism are desirable from the PsyCap (Luthans &

Youssef-Morgan) and cross-cultural PsyCap (Dollwet & Reichard, 2014) lens. The

context-specific nature of the psychological resource means that an individual could lean

heavily towards an optimistic explanatory style in certain scenarios while leaning heavily

towards a pessimistic explanatory style in other situations.

In cross-cultural PsyCap, cross-cultural optimism incorporates both the

expectancy and explanatory viewpoints (Dollwet & Reichard, 2014; Reichard et al.,

2014). Aligning with the expectancy view, individuals who are high in cross-cultural

optimism “will expect the best when interacting with people from different cultures and

have a positive outlook on future transactions” (Dollwet & Reichard, 2014, p. 1672).

Aligning with the explanatory view, those same individuals with high cross-cultural

optimism are likely to believe that positive intercultural interactions resulted from their

skills and abilities. In contrast, negative interactions result from something external and

out of their control (Dollwet & Reichard, 2014). While, to a large degree, existing

literature has not yet investigated the individual relationships between cross-cultural

optimism and the various positive organizational outcomes; cross-cultural optimism as a

part of the overall cross-cultural PsyCap construct has a positive relationship with

employee work engagement and well-being (Kotze & Massyn, 2019), cultural

intelligence and positive emotions (Reichard et al., 2014), and negative relationships with

burnout (Kotze & Massyn, 2014), and ethnocentrism (Reichard et al., 2014). As a

primary component of cross-cultural PsyCap, cross-cultural optimism also has a role in

mediating the relationship between multicultural personality traits and perceived service

quality (Maslakci & Sesen, 2019). Similarly, few studies on workplace PsyCap have

investigated the direct relationships between optimism as a second-order construct and

positive organizational outcomes. The studies that evaluated relationships at the

component level have identified positive relationships between optimism, work

engagement (i.e., vigor, dedication, and absorption), OC (i.e., affective, continuance, and

NC; Simons & Buitendach, 2013). Several studies investigating optimism embedded

within the workplace PsyCap construct have identified a positive relationship between

workplace PsyCap and OC (Hsing-Ming et al., 2017; Hussain & Nawaz, 2019; Sen et al.,

2017; Surucu et al., 2020; Yildiz, 2018), with only Idris and Manganaro (2017)

identifying no significant relationship between the constructs.

Organizational Commitment

There is a long history of scholarly research on OC, including initial research into

OC as a unidimensional construct and, more recently, using a multi-dimensional model

(Allen & Meyer, 1996). Interest in OC continued to grow throughout the last few

decades, prompting Mowday (1998) to conclude that although some research may have a

recency effect, OC remained an ever-growing construct of interest for researchers. A

meta-analysis confirmed the continued interest observing an increase from 29 relevant

articles in the 1970s to 186 in the 1990s (Mowday, 1998). Much of the research on OC

can at least partially trace roots back to the development of the organizational

commitment questionnaire (OCQ; Mowday et al., 1979). However, research dates to the

1950s that incorporates certain unidimensional factors of OC, including faithfulness and

commitment (Yildiz, 2018). The OCQ included 15 items in a 7-point Likert scale ranging

from strongly agree to strongly disagree and was intended to assess the identification and

involvement with an organization based on an individual’s belief in organizational goals,

a willingness to exert effort, and a desire to remain with the organization (Mowday et al.,

1979). Whereas researchers believed the commitment construct encompassed the three

components, they still considered the construct unidimensional; the OCQ produced a

single score and all items loaded on a single factor (Mowday, 1998). Both the definition

of OC presented by Mowday et al. (1979) and the OCQ primarily focused on attitudinal

commitment, where an individual identifies with and attaches themselves to an

organization. However, OC and the OCQ also incorporated the concept of behavioral

commitment, which researchers often defined as “overt manifestations of commitment”

(p. 225).

After Mowday et al. ‘s (1979) development of the OCQ, Meyer and Allen (1991)

acknowledged that while there were several different conceptualizations for OC, the

various concepts reflected three distinct themes, including affective attachment,

obligation to remain, and the perceived costs of leaving. This assertion led Meyer and

Allen (1991) to propose a three-component model for OC identifying AC, CC, and NC as

distinct components of commitment, rather than different types of commitment (Allen &

Meyer, 1996; Meyer and Allen, 1991; Meyer et al., 2004). The purpose of identifying

three distinct components was to acknowledge that although each component effectively

binds the employee to the organization to a greater or lesser degree, each component

could develop out of different antecedents and result in vastly different behaviors (Meyer

& Allen, 1991; Meyer et al., 2004). Meyer and Allen’s (1991) conceptualization

considers the possibility that an individual may have different levels of commitment

under each component. Each component could be higher or lower depending on internal

and external factors, suggesting that OC is somewhat state-like, malleable, and flexible to

development over time.

While studies widely use the Meyer and Allen (1991) conceptualization of a

three-component model of OC, some researchers question the model’s applicability in

light of empirical findings. Solinger et al. (2008) argued that while the three-component

model of OC has dominated the literature on OC, a more applicable model would be

Eagly and Chaiken’s (1993) attitude-behavior model. Solinger et al. (2008) based their

argument on a critique suggesting that AC describes an attitude towards the organization

within the three-component model. In contrast, NC and CC actually describe behaviors

(i.e., staying with an organization based on evaluating the perceived costs of leaving).

Furthermore, Solinger et al. (2008) argued that the three-component model of OC

proposed by Meyer and Allen (1991) is a model for predicting turnover rather than actual

OC from an attitudinal perspective. Mercurio (2015) further argued that the focus should

be on AC as the primary component of OC, suggesting that “the stream of organizational

commitment literature remains confounding, fragmented and difficult to access” (p. 391).

Mercurio’s (2015) meta-analysis of the available literature on OC suggested that most

researchers agreed on an affective, emotional, or attitudinal core to OC. Furthermore,

significant correlations exist between AC, organizational citizenship behaviors, turnover,

absenteeism, and stress (Meyer et al., 2002; Mowday et al., 1979; Solinger et al., 2008).

CC rarely correlates with such variables, and NC often correlates so strongly with AC

that it can be hard to separate the components from an empirical perspective (Ko et al.,

1997; Mercurio, 2015; Solinger et al., 2008).

Despite critiques from other researchers, the Meyer and Allen (1991) model of

OC endures and remains a very popular construct with organizational researchers

(Mercurio, 2015; Solinger et al., 2008). The popularity of the three-component model is

evidenced in PsyCap research by the significant number of researchers who have used the

model in their articles (Babalan et al., 2018; Peng et al., 2013; Pillay et al., 2014; Simons

& Buitendach, 2013; Surucu et al., 2020; Wu & Chen, 2018; Xu et al., 2020; Yildiz,

2018). However, some existing studies related to PsyCap use only a single component of

OC, primarily AC (Luthans et al., 2007b), or a one-dimensional approach (Rego et al.,

2016; Tang et al., 2019). Overwhelmingly, the studies have shown positive relationships

between PsyCap and OC as a three-component model and PsyCap and AC as a single

dimension of OC.

Affective Commitment

The development of the three-component model of OC (Meyer & Allen, 1991)

came out of a belief that previous conceptualizations of OC captured three distinct

themes. Including emotional attachment to an organization, perceived costs in leaving an

organization, and a sense of obligation (Allen & Meyer, 1996; Meyer & Allen, 1991;

Meyer & Herscovitch, 2001; Meyer et al., 2002). Within these distinct themes, the theme

relating to the emotional attachment between an individual and an organization is AC.

Researchers defined AC as “the employee’s emotional attachment to, identification with,

and involvement in the organization” (Meyer & Allen, 1991, p. 67). The description of

AC by Meyer & Allen (1991) revealed a construct rooted in Kanter’s (1968) attitudinal

commitment theory and is characterized by an employee’s desire to remain with an

organization based on their attachment, identification, and involvement (Meyer &

Herscovitch, 2001). In this context, the employee’s emotional attachment plays a

significant influence; the employee does not particularly need to stay with an

organization; rather, they want to remain with the organization (Allen & Meyer, 1996)

based on a positive evaluation of several antecedent factors. Antecedents generally

proposed in AC research primarily fall under personal characteristics, work experiences,

role or job-related characteristics, and organizational or structural characteristics (Meyer

& Allen, 1991; Mowday, 1998). While there could be numerous individual experiences

and attributes within each of these categories, it is essential to recognize that employees

evaluate whether these experiences and characteristics provide personal satisfaction and

the extent to which the factors align with their values (Meyer & Allen, 1991). The

resulting positive or negative evaluation thus influences the strength of their emotional

attachment (i.e., AC).

AC has been called the “core essence” (Mercurio, 2015, p. 391), or the “core

concept” (Solinger et al., 2008, p. 72) of OC, revealing the importance of AC for the OC

construct. However, several researchers base their critiques of the model on these

descriptions of AC. One argument is that AC and NC are strongly correlated (Meyer et

al., 2002; Solinger et al., 2008), to the point where, from an empirical perspective, the

components are difficult to differentiate from each other (Ko et al., 1997; Solinger et al.,

2008). These critiques and others have led some researchers to advocate for a

reconceptualization of OC as a construct consisting of an attitudinal component (i.e., AC)

and a behavioral component (Solinger et al., 2008). This reconceptualization considers

NC and CC as “an attitude toward a specific behavior (i.e., staying)” (p. 74). To address

some of the critiques, Meyer and Herscovitch (2001) engaged in the development of a

“general model of workplace commitment” (p. 317), in which AC was re-conceptualized

as belonging to a mindset characterized by desire. However, the previous definition for

AC remained (i.e., emotional attachment) remained. Of note, Meyer and Herscovitch

identified that the target of the emotional attachment component of AC could include the

organization, an occupation, a supervisor, but could also include much more, such as an

outcome to a course of action (e.g., continued employment). While Meyer and

Herscovitch initially suggested the “sense of being bound to a course of action of

relevance to a particular target” (p. 317) as the core essence of their general model of

workplace commitment; they acknowledged, first that such an essence would be

challenging to measure, and second, that rarely do pure forms of commitment exist and as

such it is preferential to focus on the development of AC. The reasoning behind the latter

acknowledgment is to identify that NC and CC often have very narrow focal behaviors

(i.e., staying with an organization out of obligations or perceived costs), whereas, with

AC, the focal behavior is much broader (Meyer & Herscovitch, 2001). The sense of being

bound to an action comes from the individual’s emotional attachment to the target (e.g.,

the organization). There is a high likelihood that the individual would engage in other

behaviors (i.e., discretionary behaviors), which are still of value to the organization or

target, although separate from the target behavior. Based on Meyer and Herscovitch’s

work, the potential for discretionary behaviors suggests that AC is preferable to NC or

CC, as an individual with high levels of AC is more likely to be engaged and go above

and beyond in the course of their work-related duties.

Whereas research on AC has highlighted the construct as being a preferable

component of commitment due to its strong positive correlation with positive work

outcomes (Meyer et al., 2002), research has also highlighted a correlation between

PsyCap and AC (Avey et al., 2011; Gurbuz & Yildirim, 2019; Luthans et al., 2008).

Through conducting a meta-analysis, Avey et al. (2011) identified a relationship as “the

organization (as a referent) fulfills needs for efficacy and accomplishment for those high

in PsyCap” (p. 132), which supports the employee’s identification and attachment (i.e.,

AC) towards the organization. Gurbuz and Yildirim (2019) broke this down further by

investigating the relationship between the individual personal psychological resources

and AC, identifying the most substantial predictive relationship between optimism and

AC. Gurbuz and Yildirim suggested the results reflected that a favorable opinion towards

the organization and the future, along with the other psychological resources, “will

provide the necessary fuel for motivational drive” (p. 66), resulting in higher levels of

(affective) commitment.

Normative Commitment

A sense of obligation towards an organization often characterizes NC (Allen &

Meyer, 1996; Meyer & Allen, 1991; Meyer & Herscovitch, 2001; Meyer et al., 2002).

However, this sense of obligation may reflect feelings that an employee ought to remain

with an organization due to specific pressures (e.g., family or cultural) or prior

investments (e.g., training costs, tuition reimbursement) made by the organization (Allen

& Meyer, 1996; Meyer & Allen, 1991). Meyer and Herscovitch (2001) extended the

sense of obligation to other types of targets beyond the organization in their

conceptualization of a general model of workplace commitment. In their model, NC is

used to explain commitment which falls under the obligation mindset, “NC is

characterized by the mindset that one has an obligation to pursue a course of action of

relevance to a target” (Meyer & Herscovitch, 2001, p. 316), as such a target could take

many forms such the organization, a manager, or even a work project. In this context, the

course of action may be better understood as a behavior, for example, staying with the

organization, pleasing the manager, or completing the work project. However, somewhat

different from AC, the associated behaviors would likely remain relatively narrow and

specific to the target, meaning there would be a lower likelihood of exhibiting

discretionary behaviors of value or relevance to the target (Meyer & Herscovitch, 2001).

Similar to AC, though, employees will evaluate several antecedent factors, including

personal characteristics, family and societal expectations, and organizational investments

(Meyer et al., 2002). Evaluating such factors generates the feeling of obligation the

employee feels towards the target, particularly when it comes to organizations and

organizational investments. The sense of obligation may continue until the employee

feels they have paid back their debt (Meyer & Allen, 1991).

While NC, along with AC and CC, remains prevalent in modern literature, some

researchers present arguments, particularly about NC and CC, questioning the validity

three-component model (Ko et al., 1997; Mercurio, 2015; Solinger et al., 2008). In

particular, Ko et al. (1997) argued that there is little that distinguishes NC from the more

widely accepted AC from a conceptual and empirical viewpoint. Based on their research,

Ko et al., Bergman (2005), and later Solinger et al. (2008) argued that while empirical

differentiation between AC and NC was difficult, the issue was primarily a conceptual

one. Ko et al. highlighted that believing it is right to stay with an organization (i.e.,

obligation) cannot be separated from choosing to stay with an organization that one

identifies with and is involved in (i.e., AC). Ko et al. based their assessment of NC on

Meyer and Allen’s (1991) conceptualization of NC as being rooted in antecedent factors.

Where the individual experiences both socialization to norms and feelings of reciprocity

for prior or future organizational investment. Meyer et al. (2002), in response to the

critiques of NC, suggested that the high correlation between NC and AC “is not unity” (p.

40). They pointed to variance in correlations due to geographic location and subsequent

modifications to the concept of NC to focus more on the sense of obligation as opposed

to the socialization of norms as evidence to the dimensionality of NC. However, while

the rebuttals provided by Meyer and colleagues have not completely satisfied their critics,

they have added scope and value to the OC construct. Critics have acknowledged that

despite perceived conceptual issues with NC, there continues to be value in the concept in

terms of the consequences of not staying with an organization, “which is paramount in a

vast number of studies on the matter” (Solinger et al., 2008, p. 76). Furthermore,

researchers acknowledge that NC is the least studied component of the three-component

model, including limited research on NC-specific antecedents, consequences, and

correlations. Researchers have suggested that the NC concept requires more empirical

research to contribute to a more robust understanding of the commitment component

(Allen & Meyer, 1996; Bergman, 2006; Meyer et al., 2002; Solinger et al., 2008).

Research that has focused on NC, particularly in health care organizations, has

shown some interesting results. Mousa and Puhakka (2019), in a study using physicians

at four Egyptian hospitals, identified statistically significant positive relationships

between responsible leadership, organizational inclusion, and NC. The researchers used

organizational inclusion and responsible leadership as indicators of the organizations’

adherence to being inclusive and supportive of individual and cultural differences and

their leaders’ ethical, socially aware, and engaging practices. Mousa and Puhakka

suggested that the positive relationship with NC “highlights that individuals (physicians

in this case) seek to balance their work behavior and attitude (OC in this case) with the

benefits (e.g., recognition, respect, non-work time, justice) granted from their employer”

(p. 218).

In other research specific to health care workers, Du et al. (2019) examined

“employee’s perception about their organizations’ commitment to develop their new

skills and competencies” (p. 3) in relation to their turnover intent. In this study, NC was

negatively correlated with the employees’ intent to leave and found to completely

mediate the relationship between the variables (Du et al., 2019). The researchers found a

significant mediating effect, particularly for employees who had received some level of

government subsidy. In contrast, the mediating effect was negligible for employees who

had not received any government assistance, leading the researchers to surmise that

government/organizational investment in the careers of health care workers was crucial to

enhancing NC and reducing turnover. Additionally, traditional views on NC have

suggested that AC provides a more robust prediction of employee behavior than NC;

however, as research has expanded beyond the North American population, research by

Du et al. and others have challenged that view. In particular, cross-cultural research with

Asian and other collectivist cultures has shown that NC has a much stronger predictive

relationship with turnover intent in those communities (Du et al., 2019; Yao & Wang,

2006). Occasionally the predictive relationship is even more significant than the

predictive relationship between AC and turnover intent (Chang et al., 2007) found in

most North American settings.

Continuance Commitment

Arguably, CC may be the most straightforward component of OC to understand.

The continuance component of commitment has been defined, relatively simply, as

“commitment based on the employee’s recognition of the costs associated with leaving

the organization” (Allen & Meyer, 1996, p. 253). The employee will decide, primarily

whether to remain with or leave the organization, based on an evaluation of alternatives.

Decades ago, Becker (1960) and Kanter (1968) described continuance-related concepts of

commitment, with Becker identifying that employees will commit themselves to certain

behaviors which arise out of the development of desirable side bets. For example, a side

bet could include a comfortable salary paid to an employee in return for their ongoing

employment; as such, an employee will commit themselves to the behavior (i.e.,

remaining with the organization) to maintain the side bet. Kanter suggested a similar

continuance-related concept where commitment occurs if profits result from engaging in

a particular activity and costs result from discontinuing the activity. According to Kanter,

an employee would be committed to an organization if they profit (e.g., salaries, benefits,

promotions) from remaining with the organization, and if those salaries, benefits, and

promotions are lost (i.e., costs) if they leave the organization. Incorporating certain

aspects of these earlier concepts, the three-component model of CC suggests that the

accumulated side bets represent something desirable to the employee (i.e., a profit),

which the employee can lose (i.e., a cost). Strong CC could reveal that the profit is highly

valued or that the alternatives (e.g., salaries and benefits at other companies) are not as

valuable. Alternatively, low CC could reveal that the profit is not as highly valued (e.g., a

low salary, poor benefits) or that the potential alternatives are more desirable.

In their discussion of antecedents to CC, Meyer and Allen (1991) suggested that

any side bets, investments, or alternatives that would increase the costs perceived by the

employee, could potentially be included as an antecedent. This concept may be best

understood when thinking about it in the context of staying with or leaving an

organization. If the costs associated with leaving the organization are too significant, the

employee will be more committed to staying with the organization and engage in

behaviors that support staying. As such, anything that contributes to the imbalance

between costs and profits could potentially be an antecedent. Similar to what they did

with AC and NC, Meyer and Herscovitch (2001) attempted to reconceptualize CC within

their general model of workplace commitment. CC represents a mindset characterized by

cost avoidance in the general model of commitment (Meyer & Herscovitch, 2001). While

the theoretical evaluation of costs and potential alternatives, or lack thereof, remains the

same as Meyer and Allen’s initial conceptualization of CC, they suggested there could be

a broader range of potential targets. However, Meyer and Herscovitch acknowledged that

the range of focal behaviors associated with the cost avoidance mindset would still be

relatively narrow. “If cost-avoidance is the only basis for commitment, the individual is

unlikely to engage in any other course of action not specified in the terms of the

commitment” (Meyer & Herscovitch, 2001, p. 319). This acknowledgement suggests

individuals with high levels of CC will engage in behaviors that support their target (i.e.,

do what it takes to remain employed). However, they are less likely to engage in other

behaviors (i.e., discretionary behaviors) that would benefit the employer or target.

As with NC, some researchers have argued that there are some conceptual

problems with CC. First, Ko et al. (1997) suggested that the concept of CC, which

included high personal sacrifices (i.e., costs) and lack of alternatives as subdimensions,

was flawed. The argument was that the lack of other options contributed to the increased

costs and, as such, would be an antecedent to CC rather than a subdimension (Ko et al.,

1997). Meyer and Herscovitch (2001) attempted to respond to this criticism with their

reconceptualization of CC and a renewed focus on the mindset of cost avoidance.

Solinger et al. (2008) further argued that CC often has a negative or no relationship with

the other components in the three-component model or with positive organizational

outcomes. Solinger et al. referred to Meyer et al.’s (2002) research which indicated near-

zero relationships between CC, organizational citizenship behaviors, absenteeism, and a

negative relationship between CC and job performance. Other researchers have achieved

similar results regarding the relationship with job performance finding negative or non-

significant relationships (Kaplan & Kaplan, 2018).

In direct conflict with some of the critiques of CC, some studies have found

significant positive results in the relationship between CC and positive organizational

outcomes. Lin and Chang (2015) identified that affective, normative, and continual (i.e.,

continuance) commitment had a significant positive relationship with organizational

citizenship behaviors, including altruistic behavior, employee voice, and loyal

responsibility in a sample of front-line nurses in southern Taiwan. Lin and Chang’s

results indicated that continual (i.e., continuance) commitment held the most substantial

relationship. In other research, Chang et al. (2007) suggested that cultural differences

played a significant role in findings related to OC. As such, individuals who belong to

high uncertainty avoidance cultures will “shun ambiguous situations and look for precise

alternatives” (Chang et al., 2007, p. 365), which may result in high levels of CC. Finally,

in a study of physicians in public institutions in Turkey, Yagar and Dokme (2019) found

levels of CC to be higher than levels of NC and AC. However, the differences were

relatively small; these results and others may indicate that culture and professional

context play a role in CC.

Summary and Transition

This chapter summarized the available research cross-cultural PsyCap, both in

terms of the development of the construct and the personal psychological resources,

which are the components of the higher-order construct. As a psychological state of

development (Luthans et al., 2015), the roots of PsyCap trace back to the domains of

POB, POS, and PP (Luthans & Youssef-Morgan, 2017). While a modest volume of

research exists related directly to PsyCap (i.e., workplace PsyCap), the context-specific

nature of the construct points towards a need for a better understanding across a range of

contexts, including extending knowledge into the domain of intercultural interactions

within multicultural workplaces.

A limited number of studies exist exploring cross-cultural PsyCap, with even

fewer focusing on the predictive nature of the first-order components of the construct.

The available studies have suggested that individuals with high cross-cultural hope are

more likely to be motivated to pursue intercultural interactions, including identifying

various pathways to achieve a successful interaction (Dollwet & Reichard, 2014;

Reichard et al., 2014). Individuals with high cross-cultural self-efficacy are more likely to

feel comfortable and confident in their ability to continually adapt to the needs of each

intercultural interaction (Dollwet & Reichard, 2014). Individuals with high cross-cultural

resilience are more likely to overcome difficulties and be successful when working and

interacting with people from different cultures due to their deployment of personal

resources (Dollwet & Reichard, 2014; Reichard et al., 2014). Finally, individuals with

high cross-cultural optimism expect the best and have a positive outlook towards future

intercultural interactions and believe that successful interactions result from their skills

and abilities (Dollwet & Reichard, 2014).

COR theory is a stress and motivation theory (Hobfoll, 1989). It explains how

personal psychological resources such as cross-cultural hope, self-efficacy, resilience,

and optimism travel together in resource caravans and work synergistically to form cross-

cultural PsyCap. Additionally, COR theory outlines four principles, suggesting first that

from a psychological perspective, resource loss is more impactful and resource gain

(Hobfoll, 2011). Second, individuals must invest resources to gain, maintain, or recover

other resources (Hobfoll, 2011). Third, resource gain becomes considerably more critical

when resource losses have already occurred (Hobfoll et al., 2018). Finally, individuals

will engage in increasingly irrational and desperate behaviors to defend resources at risk

of exhaustion or loss (Hobfoll et al., 2018). Through COR theory, organizations can act

as a resource caravan passageway that provides the societal and ecological conditions that

may foster, support, and enrich the personal resources of employees (Hobfoll, 2011).

COR theory also explains how and why employees would deploy personal psychological

resources during intercultural interactions. Individuals will deploy resources to gain or

protect what they genuinely value (Sungu et al., 2020), including desirable conditions

such as emotional attachments to the organization. Wright and Hobfoll (2004) identified

that in such a scenario, employees must possess the required personal psychological

resources (i.e., cross-cultural hope, self-efficacy, resilience, and optimism) and must have

the opportunity to deploy them (e.g., intercultural interactions within a multicultural

organization).

This chapter also summarized the three-component model of OC, including

history, a reconceptualization of the construct, ongoing critiques, and a summary of each

component. Meyer and Allen (1991) developed the three-component model of OC to

explain what they felt were three consistent themes about what binds an employee to an

organization that appeared throughout OC literature. AC is an employee’s emotional

attachment to an organization (Allen & Meyer, 1996), characterized by a sense of desire

(i.e., want to remain with the organization). CC is an employee’s commitment based on

evaluating costs and alternatives (Allen & Meyer, 1996), characterized by a sense of need

(i.e., costs of leaving the organization are too great). Finally, NC is an employee’s sense

of obligation towards an organization (Allen & Meyer, 1996), characterized by an

employee feeling that it is the right thing to do (i.e., should remain as the organization has

invested resources in developing the employee). Meyer and Herscovitch (2001) built

upon and extended the three-component model into a general model of workplace

commitment by expanding the range of focal targets to include occupations, supervisors,

work projects, and more. Additionally, they theorized that each commitment mindset

(i.e., desire, perceived cost, and obligation) binds the employee to a course of action

despite the expanded range of potential focal targets. The action may be of specific

relevance to the commitment (i.e., focal behavior) or non-specific to the commitment but

still relevant to the target (i.e., discretionary behavior; Meyer et al., 2004).

While there is a considerable amount of literature regarding PsyCap, cross-

cultural PsyCap, and OC, underscoring the importance of the constructs, the nature of

cross-cultural PsyCap’s influence on OC is still unclear. Researchers have called for

further research on the context-specific nature of cross-cultural PsyCap and the

relationship with other positive organizational outcomes such as OC (Kotze & Massyn,

2019; Maslakci & Sesen, 2019). Chapter 3 will address the research design and rationale

for this study, including focusing on the methodology, participant population, sampling

procedures, instrumentation, and data analysis.

Chapter 3: Research Method

Introduction

The purpose of this study is to examine the nature of the relationship between

cross-cultural PsyCap and OC for employees at a health care organization in Canada.

Health care organizations in Canada are well-known for being multicultural workplaces.

Studies using samples of health care aides from western Canada have indicated that

roughly 50% of the health care aides were born in countries other than Canada and spoke

native languages other than English (Estabrooks et al., 2015). Health care staff are also

often required to work in multidisciplinary teams, which requires collaborating with

numerous individuals to provide care to a patient (McTighe & Donovan, 2017), making

intercultural interactions relatively common-place occurrences. However, how these

intercultural interactions impact employees’ commitment to the organization has received

little attention.

The research design uses an online survey that includes the scales from Dollwet

and Reichard’s (2014) cross-cultural PsyCap questionnaire and the revised version of the

three-component model of employee commitment survey (Meyer et al., 1993). The data

collected were analyzed using PLS-SEM in a two-step process. In Chapter 3, I discuss

PLS-SEM in detail and summarize the research design, rationale, and methodology used

in the study. The methodology discussion includes overviews of the population and

sampling, recruitment and data collection, the instrumentation used and

operationalization of the constructs.

Research Design and Rationale

I deemed a correlational non-experimental research design appropriate for this

study as data were collected in a cross-sectional manner, and the study did not use control

groups or interventions. Rather, the data were self-reported by participants via online

surveys and were used to examine the nature of the relationship between cross-cultural

PsyCap (predictor) and OC (criterion) variables and the relationships between each of the

components of the latent variables. The preferred quantitative approach assumes that the

cross-cultural PsyCap and OC concepts are measurable and statistically analyzed through

numerical comparison and statistical inference. Similar procedures were used by Kotze

and Massyn (2019), who suggested that similar research would benefit from quantitative

analysis using larger sample sizes and multi-group analysis, and by Maslacki and Sesen

(2019) who suggested conducting additional research in different industry sectors to

increase the generalizability of the results. Such measurement and analysis would not be

possible using a qualitative method; participants would report the data based on their

lived experiences, and the results could not be generalized beyond the group that

provided the data. As such, similar data analysis would not be possible. Additionally, the

focus of the current study was to analyze the relationship between two constructs;

describing the relationship in terms of the socially and psychologically constructed points

of view of the employees (Gelo et al., 2008), while an admirable goal, was beyond the

scope of this study.

Methodology

Target Population

The participants for this study came from a health care organization in Alberta,

Canada. The organization employs over 10,000 staff in many different roles (over 12,000

unique positions as staff may hold multiple positions), including direct care nursing/front-

line care (front-line), maintenance/environmental services/nutrition services/other general

support (general support), and corporate/administrative/non-care related (administrative)

positions. I intended the study to be generalized to the average age range of the adult

working population; as such, I included only participants who indicated an age between

18 and 65 in the participant pool. However, as I performed data analysis using PLS-SEM,

the question of the ideal sample size was not always so clear (Memon et al., 2020). A

calculation using the Soper’s (2020) A-priori sample size for structural equation models

online calculator resulted in a recommended minimum sample size of 138. The A-priori

sample size calculator is considered a superior method of calculation as it takes into

consideration the number of variables within a model, both the latent (i.e., cross-cultural

PsyCap, OC) and the observable (i.e., cross-cultural hope, AC, etc.; Memon et al., 2020).

The A-priori minimum sample size used an anticipated effect size of 0.3, a desired

statistical power level of 90%, and a probability of 0.05. Two other popular methods for

calculating minimum sample sizes in SEM research include the inverse square and

gamma-exponential methods. When the path coefficient value with the minimum

absolute magnitude is unknown in advance, the inverse square method indicates a

minimum sample size of 160, whereas the gamma-exponential method indicates a

minimum sample size of 146 (Kock & Hadaya, 2016; Memon et al., 2020). Such results

are similar to the A-priori sample size calculator and align with the general suggestion

from Memon et al. (2020), which indicated an effective sample size for most PLS-SEM

research is greater than 160 and less than 300, but at times may be higher or lower,

dependent on the size of the population.

Geographically the organization is spread throughout the province with roughly

18 different facilities consisting of acute care hospitals, long-term care, supportive living,

and mixed care facilities. The province of Alberta has a little over 4.4 million residents.

While roughly 1 million residents belong to a visible minority or aboriginal population, a

census taken in 2016 indicated that approximately 21% of the population (845,220)

identified as immigrants. Studies specific to health care in the western provinces have

shown that a disproportionate number of immigrants to Alberta work in the health care

sector (Estabrooks et al., 2015). For example, research conducted with health care aides,

who would be considered front-line health care workers, indicated that roughly 50% of

this employee group were born outside of Canada. Additionally, most of these staff spoke

a native language other than Canada’s two official languages (i.e., English and French;

Estabrooks et al., 2015). These results would indicate that while Alberta is diverse in

terms of the places of origin of the immigrant population, health care organizations

specifically have recruited heavily from these various groups.

Sampling and Sampling Procedures

For this study, I anticipated using a proportionate stratified sampling method. The

current research focuses on healthcare employees; therefore, it made sense to stratify the

population into three substantial employment groups: front-line, general support, and

corporate employees. The overall sample consisted of employees from each stratified

group using proportionate stratification. I determined the sample size for each stratum

based on the proportional size of the employment group within the overall population

(Groves et al., 2009), which the organization provided. For example, should the front-line

staff stratum be 70% of the general population (i.e., 7,505 employees), the proportional

sample should consist of 70% front-line staff. Acharya et al. (2013) identified that

stratified sampling ensures an accurate representation between different groups from

which researchers can make estimations. Furthermore, as noted by Acharya et al.,

researchers may make comparisons between such groups; however, to estimate values for

the population, researchers would need to combine the results from all the groups or

strata (Groves et al., 2009), making the stratified sampling an appropriate method when

generalizing results. Unfortunately, the sample obtained from the surveys did not meet

the required percentage for the proportionate stratified sampling method. Rather than

discarding any data, I used a slightly disportionate sample for the analysis.

Recruitment, Participations, and Data Collection

This study used a population of employees from one organization. The employee

population is widely dispersed geographically across the province of Alberta. I made

initial inquiries with the chief human resources officer of the organization and held

follow-up meetings before the data collection phase of the study. As the organization is a

not-for-profit health care organization, research is an integral part of the medical

operations, resulting in an organization familiar with empirical research methods and an

established system through which researchers may partner with the organization. The

organization has an internal research ethics requirement in which the organization

partners with a local university ethics review board to conduct third-party research ethics

approvals. In addition, I agreed to and signed an internal knowledge transfer and data use

agreement before conducting any research with the organization.

The organization sent an email on my behalf, including an invitation to participate

in the study to staff (see Appendix A). The study invitation email provided a link to an

online informed consent form, to which the participants had to agree before being able to

begin the survey. Survey data were accessible only by me, and I saved the data in a

password-protected file on a USB drive. Through working with a partner organization, I

did not require access to any personally-identifying information. Rather, data were

collected through the web-based form and consisted of demographics and survey

response data. This study posed a minimal risk as defined by the 2013 Secretary’s

Advisory Committee on Human Research Protections (Department of Health and Human

Services, 2013).

Instrumentation and Operationalization of Constructs

Instrumentation

Cross-Cultural Psychological Capital Scale

Cross-cultural PsyCap was measured using Dollwet and Reichard’s (2014) Cross-

Cultural PsyCap scale. Dollwet and Reichard adapted the PCQ previously developed by

Luthans et al. (2007a). While the initial PCQ contained 24 items, Dollwet and Reichard’s

adapted version focuses on cross-cultural interactions and has 20 items. The adapted PCQ

uses a 7-point Likert scale ranging from 1-strongly disagree to 7-strongly agree and using

items such as I believe I can succeed at almost anything I set my mind to when working

across different cultures. Initial reliability analysis from the cross-cultural PsyCap scale

indicated Cronbach’s alpha of 0.79-0.91 for the subscales of cross-cultural hope, cross-

cultural self-efficacy, cross-cultural resilience, and cross-cultural optimism, with an

overall Cronbach’s alpha of 0.94 for the entire scale. Dollwet and Reichard determined

the scale’s reliability and validity through a multi-survey assessment exploring cultural

intelligence, openness to experience, ethnocentrism, and cross-cultural intelligence. The

results support the scale’s reliability and validity in cross-cultural skills and effectiveness

assessment (Dollwet & Reichard, 2014). Researchers have used the cross-cultural PsyCap

scale with samples from the United States (Dollwet & Reichard, 2014; Reichard et al.,

2014), South Africa (Dollwet & Reichard, 2014; Kotze & Massyn, 2019; Reichard et al.,

2014), and Northern Cyprus (Maslakci & Sesen, 2019). Studies using the cross-cultural

PsyCap scale require permission from the publisher; however, the publisher freely offers

the scale for thesis or dissertation purposes. The only caveat is that researchers must

request permission again if the work is going to be published.

Three-Component Model Employee Commitment Survey

OC was measured using the shortened version of Meyer and Allen’s (1991)

Three-Component Model of Commitment survey (Meyer et al., 2013). Meyer and Allen

conceptualized the original tool to evaluate the three sub-dimensions of OC: AC, NC, and

CC. The shortened version of the tool uses an 18-item scale, consisting of a 7-point

Likert scale ranging from 1-strongly disagree to 7-strongly agree, and uses items such

as This organization has a great deal of personal meaning to me. Meyer and Allen

performed reliability testing for the scale resulting in Cronbach’s alphas of 0.87 for AC,

0.75 for CC, and 0.79 for NC. Later, Allen and Meyer (1996) sought to further support

the scale’s validity by developing a nomological net that supported the previous evidence

establishing the construct validity of the three commitment scales. Over the years since

the development of the OC scale, numerous researchers have adapted it to other

languages and further confirmed the scale’s applicability across multiple contexts and

cultures. The three-component model employee commitment survey does not require

permission when used for non-commercial, academic purposes.

Operationalization of Variables

Cross-Cultural Psychological Capital

As previously discussed, cross-cultural PsyCap is the predictor variable in this

research. In the current study, I defined cross-cultural PsyCap as a state-like, context-

specific construct. A construct where the psychological resources of cross-cultural hope,

cross-cultural self-efficacy, cross-cultural resilience, and cross-cultural optimism are

applied directly to inter-cultural interactions in the workplace (Dollwet & Reichard,

2014). A high score on the cross-cultural PsyCap scale is assumed to indicate a higher

level of comfort, confidence, and ability in successfully interacting with people from

different cultural backgrounds. A low score indicates a lack of confidence, comfort, and

difficulties in successfully navigating such interactions. In terms of the psychological

resources which make of cross-cultural PsyCap, I defined cross-cultural hope as

“pursuing and meeting goals related to working with people from different cultures”

(Dollwet & Reichard, 2014, p. 1672). Cross-cultural self-efficacy reflects having

confidence in one’s ability to interact and communicate with employees from different

cultural groups (Dollwet & Reichard, 2014; Kotze & Massyn, 2019). Cross-cultural

resilience reflects employees being “able to keep up their performance despite obstacles,

such as language difficulties, cross-cultural conflict or other cross-cultural issues” (Kotze

& Massyn, 2019, p. 2). Finally, I defined cross-cultural optimism as being able to

“remain positive and motivated during cross-cultural interaction” (Kotze & Massyn,

2019, p. 2). The cross-cultural PsyCap scale includes four items specific to the cross-

cultural hope subscale, nine items specific to the cross-cultural self-efficacy subscale,

four specific to the cross-cultural optimism subscale, and three items specific to the cross-

cultural resilience subscale (Dollwet & Reichard, 2014).

Organizational Commitment

The criterion variable in this study is OC; in the context of this research, I defined

OC as a multi-dimensional construct in which “a psychological link between the

employee and his or her organization that makes it less likely that the employee will

voluntarily leave the organization” (Allen & Meyer, 1996, p. 252). A high score on the

three-component employee commitment survey is assumed to indicate a low likelihood

that the individual will voluntarily choose to leave the organization. A low score on the

survey indicates an increased possibility that the individual will leave the organization. In

terms of the individual components, I defined AC as an employee’s “identification with,

involvement in, and emotional attachment to the organization” (Allen & Meyer, 1996, p.

253), NC as an employee’s “commitment based on a sense of obligation to the

organization” (Allen & Meyer, 1996, p. 253), and CC as an employee’s “commitment

based on the employee’s recognition of the costs associated with leaving the

organization” (Allen & Meyer, 1996, p. 253). A high score in one component scale will

not be considered predictive of high scores in the other components (e.g., employees’

may have high CC but low AC). The lack of predictive relationship between AC, NC,

and CC may indicate one or more components may influence the individual’s decision to

stay with or leave the organization. However, should the attachment described by a

component change, the other two components may not have the combined strength to

influence the decision one way or another. The revised version of the three-component

employee commitment survey (Meyer et al., 2013) includes six items under the AC scale,

six items under the NC scale, and six items under the CC scale.

Data Analysis Plan

I used SmartPLS version 3.3.3 (SmartPLS) for the data analysis. To test the

hypothesis that a positive relationship exists between the latent variables cross-cultural

PsyCap and OC, I used PLS-SEM. This data analysis method allows a researcher to

evaluate the measurement model before assessing the structural model (Hair et al.,

2019a). I evaluated the indicator loadings, followed by an assessment of internal

consistency. I then performed additional assessments to identify convergent and

discriminant validity. The evaluation of the structural model included estimations of the

coefficient of determination (R2), effect size (f2), predictive relevance (Q2), and an

assessment of the path coefficients (Hair et al., 2019a). I performed in-sample and out-of-

sample tests for the predictive power of the models and multigroup moderation analysis

to identify whether a statistically significant difference exists based on the type of

employment the participants identified.

Research Questions and Hypotheses

The current study used the following research questions and hypotheses:

RQ1 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the AC component of OC in employees at a Canadian health care

organization?

H01 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

Ha1 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

RQ2 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the NC component of OC in employees at a Canadian health care

organization?

H02 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

Ha2 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

RQ3 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the CC component of OC in employees at a Canadian health care

organization?

H03 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

Ha3 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

RQ4 – Quantitative: Does Canadian health care organization employees’ type of

employment influence the relationship between their cross-cultural PsyCap and OC?

H04 – Canadian health care organization employees’ type of employment does not

influence the relationship between their cross-cultural PsyCap and OC.

Ha4 – Canadian health care organization employees’ type of employment

influences the relationship between their cross-cultural PsyCap and OC.

Threats to Validity

In SEM research, researchers measure validity through convergent and divergent

validity (Kumar & Upadhaya, 217); however, that is not to say that researchers ignore

other forms of validity. While the construct validity subtypes, convergent and divergent

validity, are critical pieces of the measurement model analysis, external and statistical

conclusion validity are also important considerations.

Construct Validity

Construct validity can be understood as an indicator that a test measures the

phenomenon that the researcher is interested in (Matthay & Glymour, 2020). Construct

validity is of particular importance in this study, as the constructs are latent variables that

are otherwise unobservable (Matthay & Glymour, 2020). Dollwet and Reichard (2014)

noted in their development of the cross-cultural PsyCap scale that fit indices and

reliability analysis indicated that the four-component model of cross-cultural PsyCap was

superior when compared to a one-factor model. Furthermore, Dollwet and Reichard

reduced nine items from their original hypothesized model. An analysis of regression

weights indicated the items were not accurately measuring the cross-cultural PsyCap

construct they were proposing. Based on this research, Dollwet and Reichard identified

the cross-cultural PsyCap scale as having appropriate levels of construct validity. Meyer

et al. (1993) also started with additional items as part of their scale (i.e., 30 total items

across three components); starting with additional items enabled the researchers to

perform item analyses to identify the items which best captured the commitment scales.

Meyer et al. also assessed antecedent and outcome variables to validate their selection of

items, including nursing program satisfaction, career-related work involvement, and

career plans. Based on their analysis, Meyer et al. deemed the revised three-component

model of employee commitment survey as having appropriate construct validity.

External Validity

Referring to Shadish, Cook and Campbell’s validity typology, Matthay and

Glymour (2020) defined external validity as “the extent to which study results can be

generalized to other units, treatments, observations made on units, and setting of study

conduct” (p. 376). Essentially, external validity is concerned with the generalizability of

the study findings. To avoid bias in the current study, I maintained the anonymity of the

participants throughout the research. Participants self-identified which employee group

they belonged to, and I used the demographic information to perform a multigroup

analysis. The study uses multigroup analysis to separate participants into each

employment group and determine any statistical differences. Statistical differences

between groups may indicate an underlying difference in participants’ cross-cultural

PsyCap and OC. These underlying differences, if significant, may suggest that individual

group results or overall results are not generalizable to other groups beyond health care

employees. The study also used inclusion criteria to ensure that results are generalizable

to the average age of the working population in Canada.

Statistical Conclusion Validity

Statistical conclusion validity is the “appropriate use of statistical methods to

assess the relationships among study variables” (Matthay & Glymour, 2020, p. 376).

Essentially, in this study, statistical conclusion validity equates to the question, does a

relationship exist between the variables or not? Threats statistical conclusion validity can

include; fishing, where data is analyzed repeatedly until the researcher finds a significant

result. Additionally, threats may consist of low statistical power, resulting in researchers

drawing an inaccurate conclusion about the relationship between the variables. Finally,

threats may include violating test assumptions, resulting in erroneous conclusions about

the size of an effect (Matthay & Glymour, 2020). The study addresses these threats and

mitigates the chances of making Type I and Type II errors by using a sufficient sample

size. In terms of SEM research, a rule of thumb is to use a sample size between 160 and

300 (Memon et al., 2020) to ensure sufficient data is analyzed to lead to accurate

conclusions. PLS-SEM also separates the analysis of the measurement and structural

model relationships. The separation allows researchers to use PLS-SEM for small and

large sample sizes (Hair et al., 2019a). In SEM, violating test assumptions such as

normality and multicollinearity may also threaten statistical conclusion validity. PLS-

SEM is particularly robust in protecting against violations of normality (Hair et al.,

2019a) and does not require the data distribution to be normal. Additionally, PLS-SEM

also benefits from a “high degree of statistical power” (p. 7), making the analysis method

particularly useful in exploratory and confirmatory research.

Ethical Procedures

I received ethics approval from the IRB at Walden University and the Human

Research Ethics Board at the University of Alberta (HREB) for the current study.

Alberta’s primary health care organizations have contracted out the ethics review and

approval for any research involving the organizations to the HREB. Beyond approval

from the IRB and the HREB, the organization’s research department must also approve

any research. The Chief Human Resources Officer and I had initial discussions to gauge

interest and had additional meetings after my chair and committee member approved the

study proposal. The HREB required that external student researchers receive approval

from their university ethics review (i.e., IRB at Walden University) before reviewing any

proposed research. The organization also required approval from the HREB before the

research department provided the final research approval. I have included the approvals

from the Walden IRB, the research department, and the HREB as Appendices G, E, and

Confidentiality

The participants and data in this study will remain confidential at all times. The

study used protocols ensuring no influence upon the participants from either the

researcher or the organization. Protocols included having the organization send out the

email invite to participate, completing the survey through Survey Monkey, and avoiding

personally-identifying information. Computer devices and USB drives used for the

research were password-protected to ensure the security of the stored data.

Informed Consent

While this study posed a minimal risk as defined by the 2013 Secretary’s

Advisory Committee on Human Research Protections (Department of Health and Human

Services, 2013), I still provided participants with a digital informed consent form. After

reading the description of the study and the informed consent; the participant must

indicate YES to whether they consent to participate in continuing with the survey. An

answer of NO sent the participant to the disqualification page. The informed consent,

obtained from Walden University (2020), includes a description of the study, the name

and role of the researcher, the procedures involved in the study, a statement regarding the

voluntary nature of the study, discussion of any expected risks, a privacy statement, and

finally contact information for the IRB.

Summary and Transition

Chapter 3 discussed the research design and rationale for this study, presented the

research questions and hypotheses, and introduced the study’s methodology. The chapter

also introduced the population, the sampling procedures, the recruitment of participants,

and the data collection method. Additionally, the chapter reviewed the instrumentation

for the study, discussed the operationalization of the constructs, discussed threats to the

validity, and described the ethical procedures which will take place. In summary, the

purpose of this quantitative study is to examine the nature of the relationship between

cross-cultural PsyCap and OC for employees at a health care organization in Canada.

Data will be collected from the employees at a health care organization in Alberta with

the organization’s assistance to ensure confidentiality. The study used digital versions of

the cross-cultural PsyCap scale and the three-component model employee commitment

survey to collect the data. After collection, I assessed the data using PLS-SEM. Chapter 4

will include a comprehensive breakdown of the data collection and a detailed discussion

about the results.

Chapter 4: Results

Introduction

The purpose of this study is to examine the relationship between cross-cultural

PsyCap and OC for staff at a health care organization in Canada. Cross-cultural PsyCap

was operationalized as a latent construct consisting of cross-cultural hope, cross-cultural

self-efficacy, cross-cultural resilience, and cross-cultural optimism and was measured

using Dollwet and Reichard’s (2014) cross-cultural PsyCap scale. OC was

operationalized as a latent construct consisting of AC, NC, and CC and was measured

using a shortened version of Meyer and Allen’s three-component model employee

commitment survey (Meyer et al., 2013). I chose the cross-cultural PsyCap scale and the

three-component model employee commitment survey for this study as previous studies

have shown both measures to be reliable and valid in cross-cultural PsyCap and OC

research. I used PLS-SEM in a two-step approach, similar to the process used by Kotze

and Massyn (2019), to analyze the relationship between cross-cultural PsyCap and

workplace psychological well-being. The two-step approach included data analysis to

confirm the internal consistency and construct validity and the proposed model and a

structural model assessment to test the proposed hypotheses (Sarstedt et al., 2017).

The four research questions and hypotheses upon which this research study was

structured include:

RQ1 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the AC component of OC in employees at a Canadian health care

organization?

H01 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

Ha1 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

RQ2 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the NC component of OC in employees at a Canadian health care

organization?

H02 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

Ha2 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

RQ3 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the CC component of OC in employees at a Canadian health care

organization?

H03 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

Ha3 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

RQ4 – Quantitative: Does Canadian health care organization employees’ type of

employment influence the relationship between their cross-cultural PsyCap and OC?

H04 – Canadian health care organization employees’ type of employment does not

influence the relationship between their cross-cultural PsyCap and OC.

Ha4 – Canadian health care organization employees’ type of employment

influences the relationship between their cross-cultural PsyCap and OC.

In this chapter, I address the data collection process, sample demographics, and

descriptive statistics of the research sample. Subsequently, I review the results of both the

measurement model evaluation and the structural model assessment (i.e., hypothesis

testing). Finally, I outline the results of the moderation analysis to address research

question number four, before summarizing the chapter and transitioning to Chapter 5.

Data Collection

The Walden University Institutional Review Board approved data collection for

this study (IRB – approval #05-28-21-0662235), the University of Alberta Health

Research Ethics Board (HREB – study ID Pro00109018), and the Covenant Health

Research Centre (CHRC – study# 20558). I sent an email invitation to staff at a Canadian

health care organization, inviting them to participate in an online survey by clicking a

link in the email. The link took participants directly to an online survey provided through

Survey Monkey. Participants had the opportunity to review the informed consent page

before indicating their agreement to participate. I received a total of 535 survey responses

to the survey. I downloaded the survey responses from the Survey Monkey website and

cleansed the data before uploading it into SmartPLS. As suggested in the current

literature, I manually deleted any surveys in which the respondents did not answer a

minimum of 85% of the items (Hair et al., 2017; Samani, 2016). In addition, four of the

items on the OC scale were reverse coded and, as such, required recoding. Items AC3,

AC4, AC5, and NC1, were recoded 7=1, 6=2, 5=3, 4=4, 3=5, 2=6, and 1=7 based on the

direction from the TCM Employee Commitment Survey Academic Users Guide 2004

(Meyer & Allen, 2004). Once uploaded to SmartPLS, mean replacement was used in

further data analysis to address any missing values. Hair et al. (2017) suggested that if

any indicators contain 5% or greater missing values, researchers should use casewise

deletion to remove the missing values. However, as the indicators in the current study had

less than 5% missing data per indicator, the existing literature suggests using mean

replacement to recode any missing values with the mean value specific to the individual

indicator. Missing values within the dataset were coded as -999, as suggested by Hair et

al. SmartPLS allows for this missing value indicator to be identified and automatically

replaces the missing values with the indicator mean when mean replacement is selected.

The data cleansing process removed 153 survey responses, leaving 382 responses (71.4%

completion rate) for further data analysis.

Timeframe

The online survey was active for participant completion from July 15, 2021, until

August 5, 2021. It is important to note that data collection occurred while the 2021

COVID-19 pandemic was still active. Health care employees have been particularly busy

during the COVID-19 pandemic. While I did not collect data to confirm this, the

lingering toll of COVID-19 on health care staff may have impacted survey response rates.

Additionally, as I collected data over a short period during the summer (July-August),

some employees may not have had the opportunity to respond before the survey closed

due to vacations and other leaves from work.

Normality

According to Hair et al. (2017), unlike covariance-based SEM, normal data

distribution is not required for PLS-SEM. However, whereas PLS-SEM does not assume

a normal distribution, researchers are still encouraged to differentiate between normal and

non-normal data distributions through identifying the measures of skewness and kurtosis

of their data (Hair et al., 2017). Similarly, Finney and DiStefano (2006) suggested that

one of the first analyses researchers should conduct is an assessment of the skewness and

kurtosis of the data as the results may guide some their future decisions. Skewness

measures the symmetry of the data, for example, whether the data skew to the left or right

of the distribution.

In contrast, kurtosis measures the data distribution peak (Hair et al., 2017). Data

skewed greater than +1 or less than -1 or peaked higher than +1 or lower than -1 are

considered non-normal. I calculated the skewness and kurtosis for the survey data and

included the results in Table 1. Results indicate that most data distributions are non-

normal. However, this result was expected as the data is ordinal and was collected

through a survey using a Likert scale. As mentioned by Hair et al. (2017), while

researchers must differentiate between normal and non-normal data, PLS-SEM is a

nonparametric statistical method and, as such, does not require satisfaction of the

normality assumption to perform data analysis.

Table 1

Normality of Data Distribution

Indicator

Excess kurtosis

Skewness

CCHOPE1

0.046

0.740

CCHOPE2

2.056

1.064

CCHOPE3

1.220

0.999

CCHOPE4

2.118

1.264

CCEFF1

1.951

1.224

CCEFF2

1.771

1.216

CCEFF3

0.451

0.757

CCEFF4

-0.197

0.665

CCEFF5

5.925

1.801

CCEFF6

4.820

1.701

CCEFF7

3.409

1.409

CCEFF8

1.736

1.113

CCEFF9

0.459

0.815

CCOPT1

0.223

0.634

CCOPT2

3.626

1.532

CCOPT3

2.909

1.399

CCOPT4

0.293

0.720

CCRES1

3.637

1.375

CCRES2

4.192

1.495

CCRES3

1.030

0.805

AC1

0.810

1.139

AC2

-0.962

0.247

AC3

-1.046

0.316

AC4

-0.966

0.397

AC5

-0.847

0.527

AC6

-0.414

0.607

CC1

-0.436

0.685

CC2

-0.956

0.381

Indicator

Excess kurtosis

Skewness

CC3

-0.924

0.401

CC4

-1.143

-0.047

CC5

-1.050

-0.190

CC6

-1.195

0.145

NC1

-1.147

0.015

NC2

-1.140

-0.031

NC3

-1.257

-0.056

NC4

-0.954

0.259

NC5

-0.623

0.457

NC6

-1.058

0.024

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

Study Results

Sample Demographics

The sample includes 382 participants from various positions within the

organization. From the sample, 232 (60.73%) participants identified holding front-line

positions, 52 (13.61%) held general support positions, 96 (25.13%) held administrative

positions, and two (0.52%) chose not to answer. In terms of age, participants ranged

between 18 and 65. Demographic responses indicate 71 (18.59%) participants were

between the ages of 18 and 34, 182 (47.64%) between 35 and 49, 128 (33.51%) between

50 and 65, and one (0.26%) participant chose not to answer. Finally, 322 (84.29%)

participants identified as female, 53 (13.87%) identified as male, and seven (1.83%)

preferred not to answer. Table 2 includes the sample demographics.

Table 2

Sample Demographics (Type of Employment, Age, and Gender)

Variable

Description

Frequency

Percent

Type of employment

Direct nursing care/front-line

232

60.73%

General support/maintenance

13.61%

Corporate/administrative

Did not answer

25.13%

0.52%

Age

18-34

18.59%

35-49

182

47.64%

50-65

Did not answer

128

33.51%

0.26%

Gender

Female

322

84.29%

Male

13.87%

Did not answer

1.83%

Descriptive Statistics

I calculated descriptive statistics for the survey responses under each of the

indicators. Table 3 includes the descriptive statistics for the indicators. Each survey

scored all indicators or items on a 7-point scale, with 1 indicating strongly agree and 7

indicating strongly disagree. The descriptive statistics revealed that overall, the mean (M)

for all the cross-cultural PsyCap scale indicators was below the midpoint, ranging from M

= 1.8 to M = 3.0, with standard deviation (SD) ranging from SD = 0.908 to SD = 1.459. In

terms of descriptive statistics for the OC scale, seven indicators (AC1, AC5, AC6, CC1,

CC2, CC3, and NC5) had an M below the midpoint ranging from M = 2.6 to M = 3.5 with

SD ranging from SD = 1.538 to SD = 1.863. The remaining 11 indicators had an M above

the midline ranging from M = 3.5 to M = 4.3, with SD ranging from SD = 1.747 to SD =

1.955.

Table 3

Descriptive Statistics

Construct

Indicator

Mean

Min

Max

Standard deviation

CCPsyCap

CCHOPE

CCHOPE1

2.7

1.000

7.000

1.457

CCHOPE2

2.4

1.000

7.000

1.113

CCHOPE3

2.6

1.000

7.000

1.180

CCHOPE4

2.6

1.000

7.000

1.241

CCEFF

CCEFF1

2.2

1.000

7.000

1.128

CCEFF2

2.4

1.000

7.000

1.221

CCEFF3

2.8

1.000

7.000

1.236

CCEFF4

3.0

1.000

7.000

1.459

CCEFF5

1.8

1.000

7.000

0.908

CCEFF6

1.9

1.000

7.000

0.959

CCEFF7

2.0

1.000

7.000

0.986

CCEFF8

2.4

1.000

7.000

1.163

CCEFF9

2.8

1.000

7.000

1.329

CCOPT

CCOPT1

2.8

1.000

7.000

1.193

CCOPT2

2.2

1.000

7.000

1.071

CCOPT3

2.2

1.000

7.000

1.093

CCOPT4

2.7

1.000

7.000

1.274

CCRES

CCRES1

2.1

1.000

7.000

0.966

CCRES2

2.1

1.000

7.000

0.982

CCRES3

2.4

1.000

7.000

1.021

OrgCommit

AC1

2.6

1.000

7.000

1.538

AC2

3.8

1.000

7.000

1.747

AC3

3.6

1.000

7.000

1.821

AC4

3.5

1.000

7.000

1.778

AC5

3.4

1.000

7.000

1.863

AC6

3.2

1.000

7.000

1.641

Construct

Indicator

Mean

Min

Max

Standard deviation

CC1

3.1

1.000

7.000

1.633

CC2

3.5

1.000

7.000

1.815

CC3

3.5

1.000

7.000

1.813

CC4

4.0

1.000

7.000

1.829

CC5

4.3

1.000

7.000

1.756

CC6

3.9

1.000

7.000

1.885

NC1

4.1

1.000

8.000

1.827

NC2

4.2

1.000

7.000

1.803

NC3

4.2

1.000

7.000

1.955

NC4

3.7

1.000

7.000

1.837

NC5

3.4

1.000

7.000

1.706

NC6

4.0

1.000

7.000

1.804

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment; CCPsyCap, cross-cultural psychological capital; OrgCommit, organizational

commitment

Model Estimation

To begin analysis in PLS-SEM, the models, both measurement and structural,

must first be estimated and can subsequently be evaluated (Sarstedt et al., 2017). Whereas

people often think of PLS-SEM as one single process, it is two distinct sets of equations.

The measurement model (outer model), which focuses on the relationship between a

construct and related indicators, and the structural model (inner model), which focuses on

the relationships that exist between constructs (Henseler et al., 2016). As a result, PLS-

SEM requires an estimation of model fit; researchers then assess the fit estimation for

quality. Hair et al. (2017) suggested that within SmartPLS, researchers completed the

model estimation through running the PLS algorithm using a path weighting scheme, 300

maximum iterations, and a specific stop criterion of 1x10-7. Table 4 includes the

estimation results in terms of standardized coefficients. Figures 5 and 6 visually represent

the measurement (inner) and structural (outer) models.

Table 4

Model Estimation (Outer and Inner Models)

Construct

CCEFF

0.097

0.061

0.014

CCHOPE

0.065

0.026

0.183

CCOPT

0.310

0.052

0.202

CCRES

-0.037

-0.012

-0.010

CCPsyCap

0.379

0.109

0.320

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment; CCPsyCap, cross-cultural psychological capital

Figure 5

Model Estimation of Lower-Order Constructs

100

Figure 6

Model Estimation of Higher-Order Constructs

Measurement Model Assessment

Sarstedt et al. (2017) mentioned that a two-stage process best serves PLS-SEM

analysis. The two-stage process includes an analysis of the measurement model and,

subsequently, an analysis of the structural model and hypothesis testing. I used the

disjoint two-stage approach in which the initial measurement model is constructed and

estimated using the initial indicators and lower-order constructs. Then the subsequent

101

structural model estimation uses the lower-order constructs as the indicators for the

higher-order construct (Sarstedt et al., 2019). As a “reflective-reflective second-order

construct” (Kotze & Massyn, 2019, p. 4), the cross-cultural PsyCap construct reflects the

four personal psychological resources. Similarly, the higher-order OC construct reflects

the three components of commitment, and both the psychological resources and

commitment components reflect their indicators. Current literature identifies that

reflective constructs require a specific set of analyses to evaluate the measurement model

within PLS-SEM. The collection of analyses includes reviews of indicator loadings,

composite reliability, outer loadings, average variance extracted, and heterotrait-

monotrait (HTMT) ratio of correlations (Hair et al., 2011; Hair et al., 2017, Hair et al.,

2019a; Sarstedt et al., 2017). The first step in assessing the measurement model is to

evaluate the indicator loadings, then assess the internal consistency and the construct

validity (Hair et al., 2011; Hair et al., 2019a; Sarstedt et al., 2017). To conduct such an

assessment, I used the SmartPLS PLS algorithm with the weighting scheme set to path,

the maximum iterations set to 300, and the stop criterion set to seven; as previously

mentioned, the mean replacement process automatically recoded all missing values.

Indicator Loadings

Indicator loadings, also referred to as outer loadings, are a measure of “an item’s

absolute contribution to its assigned construct” (Hair et al., 2017, p. 315). Before moving

forward with the measurement model assessment, I reviewed the indicator loadings to

ensure the associated latent construct explains more than 50% of the variance in the

indicator (Hair et al., 2019a; Sarstedt et al., 2017). Indicator loadings should be above

102

0.70; however, researchers may retain indicator loadings between 0.40 and 0.70 if the

removal of such indicators does not increase the composite reliability of the construct

above the suggested minimum. Researchers should always remove indicators with

loadings below 0.40 (Hair et al., 2011). Table 5 includes the initial indicator loadings.

Indicators CC4, CC5, and CC6 had values lower than 0.40; based on the direction from

Hair et al. (2011), I removed these indicators from the model. I then considered removing

other indicators with low values to improve the construct validity (Hair et al., 2011). I

conducted an additional validity analysis for any indicators falling below the threshold,

resulting in indicator NC1 being removed from the model due to substantial loading on

the AC construct. Table 6 shows indicator loadings after the removal of the four

indicators. While some indicator loadings still fell between 0.40 and 0.70 (e.g., CC1,

CCEFF3, CCEFF4, CCEFF9, and CCHOPE4), a review of cross-loadings indicated these

indicators were loading substantially on the appropriate constructs, and thus the

indicators were retained.

Table 5

Indicator Loadings (Prior to Removal of Indicators)

Indicator

CCEFF

CCHOPE

CCOPT

CCRES

AC1

0.818

AC2

0.725

AC3

0.770

AC4

0.783

AC5

0.787

AC6

0.826

CC1

0.640

CC2

0.765

CC3

0.681

CC4

0.128

CC5

-0.128

CC6

0.003

CCEFF1

0.744

103

Indicator

CCEFF

CCHOPE

CCOPT

CCRES

CCEFF2

0.774

CCEFF3

0.689

CCEFF4

0.576

CCEFF5

0.796

CCEFF6

0.789

CCEFF7

0.795

CCEFF8

0.729

CCEFF9

0.671

CCHOPE1

0.810

CCHOPE2

0.840

CCHOPE3

0.823

CCHOPE4

0.623

CCOPT1

0.797

CCOPT2

0.775

CCOPT3

0.855

CCOPT4

0.768

CCRES1

0.902

CCRES2

0.905

CCRES3

0.868

NC1

0.555

NC2

0.735

NC3

0.778

NC4

0.840

NC5

0.793

NC6

0.799

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

Table 6

Indicator Loadings (After Removal of Indicators)

Indicator

CCEFF

CCHOPE

CCOPT

CCRES

AC1

0.818

AC2

0.726

AC3

0.770

AC4

0.783

AC5

0.786

AC6

0.827

CC1

0.696

CC2

0.862

CC3

0.850

CCEFF1

0.742

CCEFF2

0.775

CCEFF3

0.693

CCEFF4

0.579

CCEFF5

0.794

CCEFF6

0.786

104

Indicator

CCEFF

CCHOPE

CCOPT

CCRES

CCEFF7

0.793

CCEFF8

0.730

CCEFF9

0.674

CCHOPE1

0.816

CCHOPE2

0.841

CCHOPE3

0.828

CCHOPE4

0.610

CCOPT1

0.798

CCOPT2

0.773

CCOPT3

0.855

CCOPT4

0.769

CCRES1

0.899

CCRES2

0.903

CCRES3

0.872

NC2

0.744

NC3

0.773

NC4

0.842

NC5

0.791

NC6

0.805

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

Internal Consistency

Internal consistency is synonymous with reliability; it measures outcome

consistency (Hair et al., 2017) and is vital for replicating and generalizing study results.

To analyze the internal consistency of the measurement model, I calculated a composite

reliability (CR) value for each construct. CR values greater than 0.70 indicate an

acceptable level of internal consistency (Hair et al., 2011, Hair et al., 2019a; Sarstedt et

al., 2017). CR for the lower-order constructs ranged from 0.813 to 0.921, indicating an

acceptable internal consistency/reliability level. While CR is a measure of internal

consistency, there is some evidence that it may be too liberal and most likely represents

the upper boundary of reliability values. Whereas Cronbach’s alpha represents the more

conservative lower boundary of reliability, and true reliability of the measurement model

105

falls somewhere in-between (Hair et al., 2019a; Sarstedt et al., 2017). While Cronbach’s

alpha represents the lower boundary, values should still represent acceptable levels to

contribute to an acceptable internal consistency/reliability (Hair et al., 2011; Hair et al.,

2017, Hair et al., 2019a; Sarstedt et al., 2017). Cronbach’s alpha for the constructs ranged

from 0.726 to 0.880, consistently representing the lower boundary for each construct.

Table 7 displays the Cronbach’s alpha and composite reliability results.

Table 7

Internal Consistency Analysis (Cronbach Alpha and Composite Reliability)

Construct

Cronbach’s alpha

Composite reliability

0.880

0.906

0.726

0.847

CCEFF

0.891

0.912

CCHOPE

0.779

0.859

CCOPT

0.813

0.876

CCRES

0.874

0.921

0.852

0.893

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

Table 8 indicates CR and Cronbach’s alpha results for the higher-order cross-

cultural PsyCap and OC constructs. Whereas internal consistency results for cross-

cultural PsyCap (CR = 0.907, α = 0.863) indicated an acceptable level of construct

reliability, the Cronbach’s alpha result (CR = 0.794, α = 0.636) for OC was somewhat

low; however, still satisfactory as it represented the lower boundary of reliability, and the

reliability coefficient was above 0.70 (ρA = 0.794).

106

Table 8

Internal Consistency Analysis (Cronbach Alpha and Composite Reliability – Second-

Order Constructs)

Construct

Cronbach’s alpha

Composite reliability

Cross-cultural PsyCap

0.863

0.907

Organizational Commitment

0.636

0.794

Construct Validity

In PLS-SEM, construct validity is established through assessing the level of

convergent and discriminant validity. In this context, convergent validity is “the extent to

which a construct converges in its indicators by explaining the items’ variance” (Sarstedt

et al., 2017, p. 16). As previously mentioned, an initial review of the outer loadings (i.e.,

indicator loadings) revealed issues with four indicators resulting in their removal.

Typically, outer loadings greater than 0.70 and average variance extracted (AVE) greater

than 0.50 are used to determine convergent validity (Hair et al., 2011, Hair et al., 2019a;

Sarstedt et al., 2017).

As previously outlined, outer loadings for the lower-order indicators ranged from

0.579 to 0.903, suggesting that most items have acceptable levels of validity with only

CCHOPE4, CCEFF3, CCEFF4, CCEFF9, and CC1 falling under the 0.70 mark.

Additionally, the outer loadings for the higher-order indicators ranged from 0.402 to

0.895, with only CC falling below the 0.70 level. I reviewed AVE for each latent

construct (higher and lower-order) to assess whether I should remove any additional

indicators. AVE ranged between 0.537 and 0.795 for the lower-order constructs and

between 0.586 and 0.710 for the higher-order constructs indicating the proposed

107

measurement model had acceptable convergent validity. Based on the acceptable

convergent validity, I retained all the remaining items. Table 9 displays the results for the

outer loadings and AVE.

Table 9

Convergent Validity Analysis (Factor Loadings and Average Variance Extracted)

Construct

Variable

Indicator

Outer loadings

AVE

Cross-cultural

PsyCap

0.710

CCHOPE

0.768

0.617

CCHOPE1

0.816

CCHOPE2

0.841

CCHOPE3

0.828

CCHOPE4

0.610

CCEFF

0.888

0.537

CCEFF1

0.742

CCEFF2

0.775

CCEFF3

0.693

CCEFF4

0.579

CCEFF5

0.794

CCEFF6

0.786

CCEFF7

0.793

CCEFF8

0.730

CCEFF9

0.674

CCRES

0.842

0.795

CCRES1

0.899

CCRES2

0.903

CCRES3

0.872

CCOPT

0.866

0.639

CCOPT1

0.798

CCOPT2

0.773

CCOPT3

0.855

CCOPT4

0.769

Organizational

Commitment

0.586

0.892

0.617

AC1

0.818

AC2

0.726

AC3

0.770

AC4

0.783

AC5

0.786

AC6

0.827

0.895

0.627

NC1

NC2

0.744

NC3

0.773

NC4

0.842

108

Construct

Variable

Indicator

Outer loadings

AVE

NC5

0.791

NC6

0.805

0.402

0.650

CC1

0.696

CC2

0.862

CC3

0.850

CC4

CC5

CC6

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment; blank (-), indicates the indicator was removed from analysis

The final step in the measurement model assessment is to evaluate the

discriminant validity of the constructs in the model. Discriminant validity “is the extent to

which a construct is truly distinct from other constructs, in terms of how much it

correlates with other constructs, as well as how much indicators represent only a single

construct” (Hair et al., 2017, p. 307). While researchers traditionally used the Fornell-

Larcker criterion and cross-loadings to determine discriminant validity, these measures

do not have an appropriate level of sensitivity in detecting when issues with validity are

present (Hair et al., 2017; Henseler et al., 2015). Instead, a more reliable alternative, the

HTMT ratio of correlations, was proposed by Henseler et al. (2015) and is now used

more commonly to evaluate discriminant validity. The HTMT ratio of correlations is “the

mean value of the item correlations across constructs relative to the (geometric) mean of

the average correlations for the items measuring the same construct” (Hair et al., 2019, p.

9). When constructs are conceptually distinct, the HTMT criterion should not exceed 0.85

(Hair et al., 2019a; Sarstedt et al., 2017). As seen in Table 10, the HTMT ratio of

109

correlations ranged from 0.090 to 0.838, confirming the discriminant validity of the

lower-order constructs in the measurement model.

Table 10

Discriminant Validity Analysis (Heterotrait-Monotrait Ratio of Correlations)

Construct

CCEFF

CCHOPE

CCOPT

CCRES

0.204

CCEFF

0.313

0.132

CCHOPE

0.285

0.129

0.725

CCOPT

0.422

0.128

0.784

0.669

CCRES

0.275

0.090

0.838

0.587

0.773

0.689

0.424

0.278

0.355

0.348

0.239

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

The HTMT ratio of correlations between the higher-order construct cross-cultural

PsyCap and the criterion variables AC, CC, and NC are present in Table 11. The HTMT

ratio of correlations ranged from 0.136 to 0.689, further confirming the discriminant

validity of the constructs.

Table 11

Discriminant Validity Analysis (Heterotrait-Monotrait Ratio of Correlations – Higher-

Order Constructs)

Construct

CCPsyCap

0.204

CCPsyCap

0.383

0.136

0.689

0.424

0.364

CCPsyCap, cross-cultural PsyCap; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

110

With the internal consistency and construct validity of the measurement model

confirmed, I continued to the next step of the data analysis, which was the structural

model assessment.

Structural Model Assessment

Step two in the disjoint two-stage PLS-SEM analysis (Hair et al., 2019a; 2019b;

Sarstedt et al., 2017; Sarstedt et al., 2019) analyzes the phenomenon through the

structural pathways. Sarstedt et al. (2017) suggested researchers must evaluate the

collinearity between latent constructs to conduct such an analysis. Should collinearity be

acceptable, researchers should then proceed to a calculation of the R2, Q2, f2, and finally,

the path coefficients (Sarstedt et al., 2017). Whereas Hair et al. (2019a) suggested similar

analysis methods (R2, Q2, path coefficients), they also suggested that researchers conduct

a PLSpredict procedure that will assess the out-of-sample predictive power of the model.

Multicollinearity

Collinearity issues exist when two variables are highly correlated. Evaluation of

the collinearity assesses the level of correlation between the independent (predictor)

variables. I used the variance inflation factor (VIF) to conduct this evaluation (Hair et al.,

2019a). The VIF is the reciprocal of the tolerance, the variance in one variable that the

other variables cannot explain (Hair et al., 2017). With formative models, researchers

evaluate VIF during the measurement model assessment. Indicators combine to form the

latent variable and are not interchangeable as they contribute to a specific piece of the

construct (Hair et al., 2017).

111

In contrast, with reflective models, the indicators are assumed to be a sampling of

all possible indicators of the latent variable and are relatively interchangeable. As such, in

reflective models, the assessment of multicollinearity is performed on the latent

constructs in the structural model as all indicators stemming from a construct should be

highly correlated (Hair et al., 2017). VIF results greater than 5.0 indicate critical

collinearity issues. Results between 3.0 and 5.0 suggest that possible collinearity issues

exist, and finally, results of less than 3.0 indicate that high levels of correlation between

independent variables do not exist (Hair et al., 2019a; Sarstedt et al., 2017). To conduct

the initial evaluation of multicollinearity, I ran the PLS algorithm using the lower-order

psychological resources as the indicators of the cross-cultural PsyCap variable. The

algorithm used maximum iterations of 300 and a stop criterion of 7. Table 12 includes the

VIF results for each variable. The VIF for all variables fell below 3.0, indicating that a

high level of correlation between latent constructs was not present.

Table 12

Multicollinearity (Variance Inflation Factor – VIF)

Construct

CCEFF

2.914

CCHOPE

1.622

CCOPT

2.090

CCRES

2.537

CCHOPE, cross-cultural hope; CCEFF, cross-cultural self-efficacy; CCRES, cross-cultural resilience;

CCOPT, cross-cultural optimism; AC, affective commitment; NC, normative commitment; CC,

continuance commitment

Once I assessed the structural model for collinearity issues, I could conduct the

subsequent analysis to evaluate the R2, Q2, f2, and the path coefficients (Hair et al., 2019a;

112

Sarstedt et al., 2017). The R2 serves as a measure of the explanatory power of the model,

sometimes referred to as the in-sample predictive power of the model, and can range

between 0 and 1, researchers including Sarstedt et al. (2017) and Hair et al. (2019a), have

indicated that results of 0.75, 0.50, and 0.25 are substantial, moderate, and weak results. I

used the SmartPLS bootstrapping procedure using 5000 subsamples, complete

bootstrapping, and bias-corrected and accelerated confidence intervals on a two-tailed

test with a significance level of 0.05. Table 13 displays the R2 results for the model.

Results indicate cross-cultural PsyCap explains 14.4% of the variance in AC, 10.2% of

the variance in NC, and 1.2% of the variance in CC.

I also used the bootstrapping procedure to obtain the f2. The f2 is an indicator of

“the change in the R2 value when a specific exogenous construct is omitted from the

model” (Hair et al., 2017, p. 211). Typically, f2 results from 0.02 to 0.14 are considered

small, from 0.15 to 0.34 are medium, and 0.35 are large (Hair et al., 2017). Results of the

f2 analysis can be seen in Table 13, removal of the exogenous construct cross-cultural

PsyCap would have a medium effect on AC (f2 = 0.168), a small effect on NC (f2 =

0.114), and essentially no effect on CC (f2 = 0.012).

According to Sarstedt et al. (2017) and Hair et al. (2019a), another method of

assessing the predictive accuracy of the structural model is assessing the model for cross-

validated redundancy (i.e., Q2). The Q2 assessment uses a blindfolding procedure and

assesses the extent to which the structural model predicts the endogenous latent construct

indicators (Hair et al., 2011). However, as stated by Shmueli et al. (2016) and Sarstedt et

al., “the Q2 is not a measure of out-of-sample prediction, but rather combines the aspects

113

of out-of-sample prediction and in-sample explanatory power” (Hair et al., 2019a, p. 12).

Typically, Q2 scores above zero indicate a level of predictive accuracy in the model, with

values above 0.025, 0.15, and 0.35 indicating small, medium, and large predictive

relevance of the path model of the endogenous construct (Hair et al., 2019a; Sarstedt et

al., 2017). I conducted the SmartPLS blindfolding procedure using a maximum omission

distance of seven. Positive Q2 results between cross-cultural PsyCap, and the constructs

AC and NC, reflect a relatively small predictive relevance. In contrast, the results

between cross-cultural PsyCap and CC indicate an extremely low predictive relevance.

Table 13 indicates the results for Q2. Hair et al. (2017) further suggested researchers

should calculate the effect size for Q2 (q2); however, SmartPLS does not calculate q2

automatically, and as such, I calculated manually using the equation q² (= Q²_included-

Q²_excluded)/(1-Q²_included). As there is a single exogenous variable in the structural

model (i.e., cross-cultural PsyCap), I set the Q²_excluded to 0 within the calculation. The

resulting calculations are provided for AC (0.070-0)/(1-0.070) = 0.075, CC (0.006-0)/(1-

0.006) = 0.006, and for NC (0.060-0)/(1-0.060) = 0.064. Hair et al. (2017) suggested

similar to Q2, q² results of 0.02, 0.15, and 0.35 indicate small, medium, and large effect

size, meaning the q² effect size for AC and NC were small, whereas the effect size for CC

was minimal. Table 13 includes the results for q².

Table 13

Coefficients of Determination, Cross-Validated Redundancy, and Effect Size

Relationship

Adjusted R2

CCPsyCap > AC

0.144

0.141

0.168

0.070

0.075

CCPsyCap > CC

0.012

0.009

0.012

0.006

CCPsyCap > NC

0.102

0.100

0.114

0.060

0.064

114

CCPsyCap, cross-cultural psychological capital; CCHOPE, cross-cultural hope; CCEFF, cross-cultural

self-efficacy; CCRES, cross-cultural resilience; CCOPT, cross-cultural optimism; AC, affective

commitment; NC, normative commitment; CC, continuance commitment; R2, coefficient of determination;

Q2, cross-validated redundancy; f2, effect size; q2, effect size of Q2.

Researchers use an assessment of path coefficients to establish both the

significance and relevance of the PLS-SEM correlations (Hair et al., 2014). In terms of

significance, “A path coefficient is significant at the 5% probability of error level if zero

does not fall in the 95% (bias-corrected and accelerated) confidence interval” (Sarstedt et

al., 2017, p. 22). If the lower and upper bound of the path coefficient includes zero, the

coefficient is not statistically significant. Hair et al. (2011) suggested that t-values greater

than 1.65 but less than 1.96 are significant at the 0.10 level. t-values greater than 1.96 but

less than 2.58 are significant at the 0.05 level and t-values greater than 2.58 are

significant at the 0.01 level. In terms of relevance, path coefficients between 0 and +1

indicate a statistically significant positive relationship, whereas path coefficients 0 and -1

indicate a statistically significant negative relationship (Sarstedt et al., 2017).

Additionally, the strength of the relationship increases as the path coefficient gets closer

to +1 or -1 (Sarstedt et al., 2017). Table 14 includes the model path coefficients, t-values,

and bias-corrected and accelerated confidence intervals. Results in Table 14 indicate that

there is a statistically significant, positive relationship between cross-cultural PsyCap and

AC (t = 8.967, p < 0.000) and a statistically significant, positive relationship between

cross-cultural PsyCap and NC (t = 7.018, p < 0.000). While positive and significant at the

p < 0.10 level, the relationship between cross-cultural PsyCap and CC was not significant

115

at the p < 0.050 level. The lower and upper bound of the path coefficient also included

zero, indicating the relationship was not significant.

Table 14

Path Coefficient Assessment

Relationship

Path

coefficient

Standard

deviation

value

p-value

(two-

tailed)

95% BCCI

lower limit

(2.5%)

95% BCCI

upper limit

(97.5%)

CCPsyCap > AC

0.379

0.042

8.967

0.000***

0.289

0.455

CCPsyCap > CC

0.109

0.065

1.691

0.091*

-0.094

0.210

CCPsyCap > NC

0.320

0.046

7.018

0.000***

0.225

0.400

CCPsyCap, cross-cultural psychological capital; AC, affective commitment; NC, normative commitment;

CC, continuance commitment; BCCI, bias corrected confidence interval.

* p < 0.10, *** p < 0.010;

As mentioned previously, researchers use the PLSpredict procedure to evaluate

the out-of-sample predictive power of the structural model (Hair et al., 2019a; 2019b;

Shmueli et al., 2019). Unfortunately, as Shmueli et al. (2019) noted, PLSpredict is still a

relatively new procedure. While developers have incorporated PLSpredict into some

PLS-SEM software, such as SmartPLS, many researchers are still uncertain about

interpreting the PLSpredict results resulting in an inconsistently used procedure. While an

in-depth discussion about PLSpredict is beyond the scope of this study, Shmueli et al.

outline three significant features of the PLSpredict procedure. First, PLSpredict uses both

training and holdout samples, whereas the “training sample is a portion of the overall

dataset used to estimate the model parameters” (p. 2325), the holdout sample is “the

remaining part of the dataset not used for model estimation” (p. 2325). Second, the

PLSpredict procedure uses indicator values from the holdout sample and the training

116

sample model estimates to generate a prediction. Third, Shmueli et al. suggest that a

slight difference between the actual and PLSpredict values indicates a high predictive

power for the model. In contrast, a considerable variation in actual and predicted results

implies a model has lower predictive power.

Researchers can evaluate the prediction statistics using k=10-fold cross-validation

splits and r=10 prediction repetitions, as Shmueli et al. (2019) suggested. Prediction

statistics include the mean absolute error (MAE; when prediction errors are highly non-

symmetrically distributed) and root mean squared error (RMSE; when prediction errors

are highly symmetrically distributed) to assess the predictive power of the model. If the

naïve benchmark (Q2Predict) is less than zero, the model lacks predictive power; however,

if the Q2Predict is greater than zero, MAE and RMSE can be assessed. The PLSpredict

procedure also uses a linear regression model (LM) as a second benchmark and compares

the LM RMSE/MAE with the PLS-SEM RMSE/MAE. If the PLS-SEM values are less

than the LM values for none of the indicators, the model lacks any predictive power. If

PLS-SEM values are less than the LM values for a minority of the indicators, the model

has low predictive power. If PLS-SEM < LM for most of the indicators, the model has

moderate predictive power, and if PLS-SEM < LM for all indicators, the model has high

predictive power (Shmueli et al., 2019). As seen in Table 15, PLS-SEM RMSE is less

than LM RMSE for most indicators, implying that the model has a moderate, out-of-

sample predictive power.

117

Table 15

PLSpredict Procedure

Item

PLS-SEM

RMSE

MAE

Q²_predict

RMSE

MAE

Q²_predict

PLS-SEM -

RMSE

AC3

1.805

1.549

0.022

1.800

1.550

0.027

0.005

Item

PLS-SEM

RMSE

MAE

Q²_predict

RMSE

MAE

Q²_predict

PLS-SEM -

RMSE

AC5

1.845

1.561

0.025

1.841

1.559

0.029

0.004

AC6

1.556

1.266

0.105

1.557

1.252

0.104

-0.001

AC1

1.425

1.126

0.147

1.422

1.102

0.150

0.003

AC2

1.660

1.404

0.104

1.673

1.407

0.091

-0.013

AC4

1.779

1.522

0.003

1.766

1.505

0.018

0.013

CC1

1.637

1.323

0.000

1.656

1.332

-0.023

-0.019

CC3

1.814

1.539

0.004

1.829

1.550

-0.013

-0.015

CC2

1.821

1.563

0.000

1.841

1.579

-0.022

-0.020

NC5

1.652

1.342

0.056

1.664

1.346

0.042

-0.012

NC6

1.731

1.457

0.067

1.735

1.468

0.062

-0.004

NC4

1.751

1.461

0.087

1.758

1.456

0.080

-0.007

NC2

1.763

1.507

0.043

1.764

1.501

0.042

-0.001

NC3

1.911

1.649

0.040

1.924

1.662

0.026

-0.013

AC, affective commitment; NC, normative commitment; CC, continuance commitment; RMSE, root mean

squared error; MAE, mean absolute error; LM, linear model; PLS, partial least squares; SEM, structural

equation model; Q²_predict, naïve benchmark.

Moderation Analysis

I conducted a multigroup moderation analysis to determine whether the

participants’ employment type (i.e., front-line care staff, general support staff, and

administrative staff) moderates the relationship between cross-cultural PsyCap and OC.

To calculate the moderation effects of employment type on the relationship between

cross-cultural PsyCap and OC, I used a multigroup analysis (MGA). SmartPLS allows

the researcher to separate data into the defined groups; for the present study, one of the

goals was to identify if there was a significant difference in the relationship between

118

cross-cultural PsyCap and OC displayed by each employment type. Cheah et al. (2020)

suggested MGA is superior to a moderated regression for this purpose as simply using

pooled data for an assessment may not identify differences in heterogeneous groups,

whereas MGA will. The MGA function in SmartPLS allows the researcher to generate

specific data groups based on the number of unique values (i.e., indicators) contributing

to the variable. The employment type variable contained three unique values, 1) front-line

staff, 2) corporate/administrative staff, and 3) general support staff. SmartPLS generated

individual groups by employment type and computed the MGA. Table 16 includes the

path coefficients of the relationships between cross-cultural PsyCap, AC, NC, and CC for

each group. The table shows that the path coefficients for the relationship between cross-

cultural PsyCap and AC for all three employment groups were statistically significant (p

< 0.01). Path coefficients for the relationships between cross-cultural PsyCap and CC

were statistically significant (p < 0.01) for the front-line group. Path coefficients were

also statistically significant between and cross-cultural PsyCap and NC for both the front-

line (p < 0.01) and the general support groups (p < 0.01). SmartPLS provides parametric

testing results in the MGA calculation to determine whether the difference between

groups was significant. As shown in Table 17, while there were minimal differences in

the relationships (i.e., path coefficients) between cross-cultural PsyCap and AC, NC, and

CC between the three employment groups, the differences between groups did not reach

statistical significance for any of the relationships.

119

Table 16

Multi-Group Analysis Bootstrapping Results

Path

coefficients

(CA)

Path

coefficients

(FL)

Path

coefficients

(GS)

t-Value

(CA)

t-Value

(FL)

t-Value

(GS)

p-Value

(CA)

p-Value

(FL)

p-Value

(GS)

CCPsyCap -> AC

0.302

0.391

0.481

2.703

8.397

4.257

0.007***

0.000***

CCPsyCap -> CC

0.160

0.181

-0.092

1.040

2.789

0.326

0.298

0.005***

0.744

CCPsyCap -> NC

0.239

0.357

0.391

1.772

7.116

2.612

0.077*

0.000***

0.009***

CA, Corporate/Admin; FL, Front Line; GS, General Support; AC, affective commitment; NC, normative commitment; CC, continuance commitment;

CCPsyCap, cross-cultural psychological capital.

**p < 0.05, *** p < 0.01,

Table 17

Multi-Group Analysis Parametric Testing

Path

coefficients-

diff (CA vs

FL)

Path

coefficients-

diff (CA vs

GS)

Path

coefficients-

diff (FL vs

GS)

t-Value (CA

vs FL)

t-Value (CA

vs GS)

t-Value

(FL vs GS)

p-Value

(CA vs

FL)

p-Value

(CA vs

GS)

p-Value

(FL vs

GS)

CCPsyCap -> AC

-0.089

-0.179

-0.09

0.872

1.038

0.810

0.384

0.301

0.419

CCPsyCap -> CC

-0.022

0.252

0.273

0.154

0.862

1.439

0.878

0.390

0.151

CCPsyCap -> NC

-0.118

-0.153

-0.034

1.019

0.718

0.271

0.309

0.474

0.787

CA, Corporate/Admin; FL, Front Line; GS, General Support. AC, affective commitment; NC, normative commitment; CC, continuance commitment;

CCPsyCap, cross-cultural psychological capital.

120

Hypothesis Testing

RQ1 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the AC component of OC in employees at a Canadian health care

organization?

H01 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

Ha1 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the AC component of OC in employees at a Canadian health

care organization.

The path coefficient between cross-cultural PsyCap and AC was 0.379 (p < 0.01),

indicating a statistically significant positive relationship. As the path coefficient was

statistically significant and positive, I rejected the null hypothesis and accepted the

alternative hypothesis, meaning AC increases in correlation with cross-cultural PsyCap.

RQ2 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the NC component of OC in employees at a Canadian health care

organization?

H02 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

121

Ha2 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the NC component of OC in employees at a Canadian health

care organization.

The path coefficient between cross-cultural PsyCap and NC was 0.320 (p < 0.01),

indicating a statistically significant positive relationship. As the path coefficient was

statistically significant and positive, the null hypothesis was rejected, and the alternative

hypothesis was accepted, indicating NC increases in correlation with cross-cultural

PsyCap.

RQ3 – Quantitative: What is the nature of the relationship between cross-cultural

PsyCap and the CC component of OC in employees at a Canadian health care

organization?

H03 – There is no statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

Ha3 – There is a statistically significant, positive relationship between cross-

cultural PsyCap and the CC component of OC in employees at a Canadian health

care organization.

The path coefficient between cross-cultural PsyCap and CC was 0.109 (p =

0.091), indicating a statistically non-significant positive relationship. As the path

coefficient was statistically non-significant, the alternative hypothesis was rejected, and

the null hypothesis was accepted, meaning that while CC increases slightly in correlation

with cross-cultural PsyCap, the increase was not statistically significant.

122

RQ4 – Quantitative: Does Canadian health care organization employees’ type of

employment influence the relationship between their cross-cultural PsyCap and OC?

H04 – Canadian health care organization employees’ type of employment does not

influence the relationship between their cross-cultural PsyCap and OC.

Ha4 – Canadian health care organization employees’ type of employment

influences the relationship between their cross-cultural PsyCap and OC.

The MGA results indicate statistically significant positive relationships between

cross-cultural PsyCap and AC for all employment types (p < 0.01). Between cross-

cultural PsyCap and CC (p < 0.01) for the front-line group. Finally, between cross-

cultural PsyCap and NC for both the front-line and the general support groups (p < 0.01).

When comparing the path coefficients for each employment type, the resulting

differences were not statistically significant. As the MGA results are not statistically

significant, I rejected the alternative hypothesis and accepted the null hypothesis.

Meaning the type of employment did not influence the relationship between cross-

cultural PsyCap and OC.

Summary and Transition

The current study collected data via an online survey link emailed out to

employees of a local health care organization with an invitation to participate in the

research. The survey link directed interested participants to an online survey created and

hosted on Survey Monkey. The online survey link remained active for three weeks, and a

reminder invitation was emailed to employees roughly halfway through the collection

period. I removed data that did not meet the specific requirements of the analysis model

123

and recoded the remaining missing values before loading the data into SmartPLS. A

review of normality indicated that the data had a non-normal distribution; however, data

analysis continued as the analysis model did not assume a normal distribution. I

calculated sample demographics and descriptive statistics before using a PLS-SEM

disjoint two-stage analysis. First, the measurement model was analyzed, resulting in the

removal of four indicators before confirming the internal consistency/reliability,

convergent validity, and discriminant validity of the model. With the quality of the

measurement model confirmed, I assessed the structural model for path coefficients, bias-

corrected confidence intervals, R2, Q2, f2, and q2, and finally, out-of-sample predictive

power. Results from the structural model analysis indicated that statistically significant

positive relationships existed between cross-cultural PsyCap and two components of OC

(AC and NC). There was also a positive relationship with the CC component; however,

the relationship did not reach statistical significance. Finally, the current study used an

MGA analysis to identify differences in the relationships specific to each employment

group. Results of the MGA indicated that while there were slight differences in terms of

the relationships between cross-cultural PsyCap and OC, the differences did not reach the

level of significance.

Chapter 5 includes an interpretation of the statistical results identified in the

previous chapter, an acknowledgement of any study limitations, and further discussion in

terms of recommendations and implications resulting from the research. Finally, the

chapter ends with a discussion of the conclusions resulting from the study.

124

Chapter 5: Discussion, Conclusions, and Recommendations

The purpose of this quantitative correlational study was to explore the nature of

the relationship between cross-cultural PsyCap and OC for employees at a health care

organization in Canada. The nature of this study was quantitative research using a cross-

sectional, correlational design and self-reported data collected through surveys. The

online survey consisted of the Three-Component Model Employee Commitment Survey

(Meyer et al., 1993) and the Cross-Cultural Psychological Capital Scale (Dollwet &

Reichard, 2014). I sent out a survey link to all employees of a health care organization in

Canada, and 535 individuals responded. Data were reviewed and cleansed, and I included

382 responses for further statistical analysis using the PLS-SEM method.

The results from the PLS-SEM analysis indicated that statistically significant

positive relationships exist between cross-cultural PsyCap and both AC and NC. Results

also showed that while a positive relationship existed between cross-cultural PsyCap and

CC, the relationship was not statistically significant and yielded little explanatory power

or predictive accuracy. Additionally, I performed a MGA using the three employment

types identified in the study. The MGA indicated that the strength of the relationship

between variables was different in each group; however, these differences did not reach

statistical significance. The following chapter will include an interpretation of the results,

the limitations of the study, recommendations resulting from the research, implications

for social change, and a summary of conclusions.

125

Interpretations of Findings

Research on PsyCap has consistently shown positive relationships with positive

organizational outcomes regardless of the domain (Adler & Aycan, 2018; Badran &

Youssef-Morgan, 2015; Bergheim et al., 2015; Blanchet-Garneau & Pepin, 2015; Bogler

& Somech, 2019; Firestone & Anngela-Cole, 2016; Kotze & Massyn, 2019).

Additionally, most research has shown that in terms of the relationship between PsyCap

and OC, the most significant relationship is usually with AC (Avey et al., 2011; Gurbuz

& Yildirim, 2019; Luthans et al., 2008). However, until the current study, there was a gap

in the scholarly literature regarding the intercultural interactions domain and how cross-

cultural PsyCap specifically interacted with OC. Results from the present study supported

the prior findings in terms of a positive relationship with positive organizational

outcomes and the strength of the relationship with AC. Results from the present study

also supported specific findings by Gurbuz and Yildirim (2019), showing the most

significant relationship between the personal psychological resources and the components

of OC existed between optimism and AC.

The present study used the disjoint two-stage approach to analyze the hierarchical

PLS-SEM’s lower-order and higher-order constructs. Upon initial review of the model

estimation, it was clear that several indicator loadings had variances not substantially

explained by the associated latent variable requiring removal from the model. Removal of

the indicators improved the CR and AVE for both CC and NC, which confirmed the

model’s reliability and convergent validity (Hair et al., 2017). Discriminant validity of

the model was confirmed via the HTMT criterion, reflecting all the remaining indicators

126

were representing a single, distinct construct (Hair et al., 2017). Within the disjoint two-

stage approach, once the lower-order constructs are estimated, latent variables are then

used as the indicators for the higher-order construct (i.e., cross-cultural PsyCap).

Subsequently, I assessed the inner/structural model for path coefficients, bias-corrected

confidence intervals, R2, Q2, f2, q2, and finally, out-of-sample predictive power.

The results from these analyses indicated several interesting findings. First, a

statistically significant relationship exists between cross-cultural PsyCap and AC and

cross-cultural PsyCap and NC; second, a positive relationship exists between cross-

cultural PsyCap and CC; however, that relationship was not statistically significant.

Third, there was no statistically significant difference in the relationships when analyzed

by employment type. While the relationships were the focus, it is also important to note

the following. Despite showing statistically significant results, the relationship between

cross-cultural PsyCap and AC had relatively weak explanatory power (R2 = 0.144), a

moderate effect size (f2 = 0.168), and low predictive relevance (Q2 = 0.070). The

relationship with NC had weak explanatory power (R2 = 0.102), a small effect size (f2 =

0.114), and low predictive relevance (Q2 = 0.060). Finally, the relationship with CC was

essentially non-existent for each of the measures. However, the PLSpredict procedure

indicated that the overall model had moderate out-of-sample predictive power. The MGA

results indicated that while each employment group experienced the relationship between

their cross-cultural PsyCap and their OC differently, there was not a significant enough

difference in that experience to suggest that employment type moderates the relationship.

127

The current study extends the understanding of intercultural interactions within

the health care industry in Canada. Results indicate that, on average, the health care

workers were confident in their ability to interact across different cultures as the mean for

all indicators skewed to the positive with a relatively small standard deviation. However,

results showed a negative trend in OC; six indicators skewed towards the negative and

nine were positive. Additionally, whereas cross-cultural PsyCap explains only 14.4% of

the variance in AC, 10.2% variance in NC, and 1.2% variance in CC, further increases in

cross-cultural PsyCap would result in only a modest rise in OC.

The present study extends the understanding of Hobfoll’s COR theory as higher

levels of cross-cultural PsyCap were observed in health care workers in an environment

that provided them with the opportunity to use their personal skills and resources (Wright

& Hobfoll, 2004). These findings align with our understanding of COR theory and

support Sungu et al.’s (2020) research. While confirmation of whether a causal

relationship exists was beyond the scope of the current study, scholarly literature on

cross-cultural PsyCap may benefit from future research into this relationship in

comparison with environments that do not provide a similar opportunity to use the

resources and skills.

This study also addresses a gap in the scholarly literature. The present study is the

only study that evaluates how cross-cultural PsyCap influences affective, normative, and

CC and how that relationship may differ across employment groups. The study results

indicate that while cross-cultural PsyCap positively influences all three components, the

strength of that influence varies from moderate (AC) to relatively insignificant (CC).

128

Results of the present study also indicate that identified relationships do not significantly

differ across different employment groups.

Limitations of the Study

Unfortunately, proportionate stratification was not possible due to lower response

rates in front-line and general support staff groups. Front-line staff make up roughly 63%

of the employees at the organization; however, only 61% of respondents to the survey

identified as belonging to this group. Similarly, whereas 19.4% of total staff in the

organization hold general support positions, only 13.4% of respondents self-identified as

such, meaning that proportionate stratification was not possible. As respondents self-

reported their employment type through the demographic questions of the survey, the

study used the employment type variable to differentiate between employment groups

when performing the MGA.

Other study limitations include the generalizability of the results beyond the

health care environment, the research design, and the use of self-reported data. Whereas

there is a relatively large body of research on general and workplace PsyCap (Luthans &

Youssef-Morgan, 2017), cross-cultural PsyCap research remains in its infancy. While it is

generally assumed that the various domains of PsyCap will follow similar patterns in

terms of findings as general PsyCap, it remains to be proven true in all instances. The

present study used participants from a single health care organization with facilities in a

single province in Canada. Specific research has shown that the environment and cultural

context can vary the expected results (Chang et al., 2007; Vandenberghe, 2003). There is

129

a limitation where researchers may find different results leading to different conclusions

when using participants within a different cultural context, workplace, or environment.

A second limitation to the present study is the quantitative analysis of the data.

While robust and informative from a correlational perspective, the quantitative analysis

does not allow for a deeper understanding of why health care workers experience cross-

cultural PsyCap in the reported manner. Additionally, the quantitative analysis does not

explain why OC does not appear to be impacted to the same degree as seen in previous

research on general PsyCap and OC. While appropriate for exploratory study and

particularly robust for such data analysis (Hair et al., 2019a), PLS-SEM can only estimate

the relationships in a model and is therefore limited in its ability to explain the

correlation. A final related but distinct limitation is the use of self-reported data. The

current study used an email invitation linked to an online survey to collect data from

participants. Despite efforts to maintain the confidentiality of the participants, there is

potential that participants were still subject to social desirability bias. Social desirability

bias may have existed due to the survey questioning participants’ beliefs and attitudes

towards situations and interactions involving culturally diverse individuals and their

commitment to their employer. The self-reported data via online survey may also present

issues with common method bias due to collecting data from a single source (Bogler &

Somech, 2019), collecting data in sequential order (Munyaka et al., 2017), or collecting

data on multiple constructs with the same composite survey (Xu et al., 2017).

130

Recommendations

While this research addressed a gap in understanding how cross-cultural PsyCap

relates to positive organizational outcomes, future research has many avenues. The

current study evaluated the relationship between cross-cultural PsyCap and OC

specifically. As the literature on cross-cultural PsyCap is still relatively scant, future

research may consider extending this to other positive organizational outcomes such as

job satisfaction, job involvement, organizational citizenship behaviors, and performance

(Kotze & Massyn, 2019). Additionally, while the current study extended the concept of

cross-cultural PsyCap into the health care industry in Canada, future research may extend

similar lines of questioning into other contexts (Maslakci & Sesen, 2019). Contexts for

prospective study may include different industries, public and private entities, and other

cultural and environmental contexts. Future researchers may also pursue research designs

and other inquisitive methods to better understand the qualitative nature of cross-cultural

PsyCap. A deeper understanding may include investigating how it interacts with different

constructs and what those relationships and constructs signify to individuals. Finally, the

scholarly literature would benefit from future research which uses research designs (i.e.,

longitudinal) and data collection methods (i.e., multi-source feedback) that explore

causality (Maslakci & Sesen, 2019) and reduce common method bias.

Implications

Intercultural interactions within a workplace can be significantly stressful and

psychologically draining for employees when they do not have the confidence,

motivation, and personal resources to interact across cultures effectively (Kotze &

131

Massyn, 2019). The current study investigated the nature of the relationship between a

combination of synergistic personal resources, namely cross-cultural PsyCap and OC,

which previous research has associated with turnover intent (Sen et al., 2017; Seo &

Chung, 2019) and workforce engagement (Basit, 2018; Peng et al., 2013). This study fills

a gap in the literature regarding the relationship between these variables and can provide

several original contributions to the scholarly literature. Whereas previous literature on

workplace PsyCap has shown a statistically significant positive relationship with OC, this

study supported those earlier findings by extending the research into a new and scarcely

researched domain of intercultural interactions. The study further supported the

similarities across the various domains of PsyCap, as the observed results aligned with

the prior research showing a dominant relationship between different domains of PsyCap

and AC compared to the other components of OC.

This study may also provide an original contribution to the scholarly literature on

OC. Although the three-component model of OC is well-researched, the impact of cross-

cultural PsyCap on OC was not well-understood until the current study. The results

indicated that while statistically significant positive relationships existed with AC and

NC, the real-world impact of those relationships was relatively minimal and did not

reflect similar levels of impact to prior research on workplace PsyCap and OC.

Additionally, the current study added to the scholarly literature by outlining that there

was essentially no relationship between cross-cultural PsyCap and CC for this group of

participants. CC represents the employees’ evaluation of the costs associated with leaving

an organization. This study’s results indicate that employee confidence and motivation

132

towards their ability to interact interculturally may not figure highly into their cost-

avoidance analysis.

Finally, this study may also provide a basis for several positive social change

implications. The findings indicated that cross-cultural PsyCap, like PsyCap in other

domains, is correlated with positive organizational outcomes such as OC. These results,

similar to research on employee well-being (Kotze & Massyn, 2019), cultural

intelligence, cross-cultural adjustment, ethnocentrism, openness to experience (Dollwet

& Reichard, 2014), multicultural personality traits, and perceived service quality

(Maslakci & Sesen, 2019), reflect the PP underpinnings of cross-cultural PsyCap.

Whereas PP focuses on the positive aspects of life (Seligman, 2019), the results show that

improvements in employee cross-cultural PsyCap correlate with positive increases in

employee personal resources and the overall outcomes for the organizations. The positive

focus towards intercultural interactions adds additional importance as workplaces

continue to grow and advance in cultural diversity. This study may also contribute to

positive social change by providing a quantitative basis for organizations to allocate

valuable learning and development resources. Organizations that invest in their

employees want to know what kind of return on their investment they can expect, both in

terms of financial return and other positive outcomes. This study provides organizations

with a better understanding of the positive organizational outcomes obtained with

increased investment in the development of cross-cultural PsyCap. Prior research has

shown that relatively short training sessions can enhance cross-cultural PsyCap (Dollwet

& Reichard, 2014; Kotze & Massyn, 2019). While the overall impact of increased cross-

133

cultural PsyCap on OC may be somewhat limited, the findings contribute to a body of

research indicating that short bursts of training result in several positive individual and

organizational outcomes.

Conclusions

Cultural diversity is growing in many organizations. If employees do not have the

confidence, motivation, and personal resources to work in such environments effectively,

the employees and the organization could experience significant adverse impacts. This

quantitative study aimed to explore the nature of the relationship between cross-cultural

PsyCap and OC in health care employees in Canada. I developed a model to examine this

relationship between two hierarchical constructs, and after estimation, I found the model

to be appropriate for further analysis. I used the disjoint two-step procedure, including

measurement and structural models, to assess the quality of data and the significance of

the relationships. My assessment found statistically significant, positive relationships

between cross-cultural PsyCap and affective and NC. The study also found a connection

between cross-cultural PsyCap and CC; however, the relationship was minimal and not

significant. In addition, this study found that despite working in different positions, the

health care employees did not experience substantial variance in their relationship

between their cross-cultural PsyCap and OC based on their employment type.

This study fills a gap in the scholarly literature on these constructs and concepts.

Through fill that gap, this study supports the currently available research on cross-

cultural PsyCap, personal psychological resources, and the three-component model of

OC. It provides original contributions to understanding cross-cultural PsyCap, OC, and

134

positive social change. Additionally, future research will be needed to understand the

qualitative context in which these relationships exist and continue to add to a fuller

understanding of cross-cultural PsyCap. Organizations that view their workforce as a

source of competitive advantage may interpret the findings from this study, along with

other research on cross-cultural PsyCap, and use this evidence to help inform decisions

on valuable resource allocations to improve outcomes for the workforce and the

organization.

The Relationship Between Cross-Cultural Psychological Capital and Organizational Commitment.

Students also viewed