The important role of work values:
The Moderating Effect of Work Values on the Relationship between
Organizational Justice and Organizational Citizenship Behavior
Master Thesis
August, 2014
Student: Lina/ Student Number: 10609431
University of Amsterdam, Faculty of Economics and Business
Business studies specialized in Strategy track
Supervisor: Dr. Katinka Quintelier
University of Amsterdam, Amsterdam Business School
Abstract
Individual behavior, in particular organizational citizenship behavior (OCB) has been attracting attention in the field of behavioral strategy field since 1983. Existing literature suggests that organizational justice is a significant predictor of OCB. However, individual differences are neglected in discussing the factors influencing OCB. This paper focuses on whether the effect of organizational justice on OCB is moderated by four work values that are regarded as individual differences, namely intrinsic, extrinsic, social, and prestige work values. In this study, a sample size of 250 is used to test hypotheses. The results indicate that there are significant moderating effects of intrinsic work values and social work values on the relationship between organizational justice and OCB; while extrinsic work values and prestige work values do not affect the relationship. Based on this research, intrinsic and social work values can directly affect OCB and enhance the positive influence of organizational justice on OCB. This paper contributes to both theory and practice.
Table of Contents
ABSTRACT ……… 2
I Introduction……… 5
II Literature review………..………… 8
1. Organizational Citizenship Behavior and Organizational Justice…… 8
2. Work values, organizational justice and organizational citizenship behavior……… 15
3. Literature gap and research question………..… 18
III Methodology……….……… 21 1. Sample……… 21 2. Procedure……… 21 3. Independent variables……… 22 4. Dependent variables……… 24 5. Mediating variable……….. 25 6. Moderating variables……… 26 7. Control variables……… 28 IV Results……… 29 1. Data Cleaning……… … 29 2. Descriptive Statistics……… 30
3. Correlation Analysis and Reliability Analysis ……… 33
4. Normality Analysis……… 38
V Discussion………... 55
1. Main findings……… 55
2. Additional findings……….. 57
3. Implications……….. 57
4. Limitations and future research……… 58
VI Conclusion……….. 59
References………. 61
Appendix A: Questionnaire……… 66
Ivitation letter……… 66
I Introduction
“The stellar universe is not so difficult of comprehension as the real actions of other people” said Marcel Proust. This means that behavior of people is too hard to fully understand its mechanism (Robbins, 2012, p.3). Specifically, behavior in the work place is diverse because of different thoughts. For managers, knowing the nature of behavior and encouraging positive behavior are significant to enhance a firm’s performance. The importance of employees’ behavior has slowly been realized in improving a firm’s performance, and individual behavior has started to gain extensive attention in the field of organizational behavior.
Resources, capabilities and structures have long been major corporate-level research topics in the field of strategic organization (Barney, 1986; Barney 1991; Peteraf 1993; Dierickx & Cool, 1989; Grant, 1996; Teece et al., 1997; Felin & Foss, 2005). Previous literature has focused on how firms obtain resources, build various capabilities, and establish different structures. Recently, however, scholars have tended to focus on analyzing how people can influence a firm’s performance at the individual level, through motivation and behavior of employees (Judge & Bretz, 1992; Ros, Schwartz, & Surkiss, 1999; Gahan & Abeysekera, 2009). This shows that scholars recognize the effect of individual behavior on improving a firm’s performance, because resource, capabilities and structures cannot be used without human resources (Lee and Allen, 2002; Holtz and Harold, 2013).
It is interesting to examine individual behavior and, in particular, organizational
citizenship behavior (OCB), because OCB contributes to organizations by improving
Bateman & Organ, 1983; Farh, Podsakoff, & Organ, 1990). OCB was created by Organ and his colleagues, and they define OCB as the individual behavior which is not motivated by any reward system. It is also defined as an extra-role behavior that is beyond the job description (Bateman and Organ, 1983; Smith, Organ and Near, 1983).
Studies by Posdakoff and Mackenzie (1994) and Podsakoff, Whiting, Podsakoff, & Blume (2009) found that OCB is positively related to different indices of a firm’s performance, including productivity, efficiency, costs, and profitability. Bateman and Organ (1983) define organizational citizenship behavior as the behavior and attitudes that are beneficial to the organization, and as individual behavior that is not restricted by organizational rules. Most research has already investigated the potential antecedents of OCB, such as employee attitudes (Bateman & Organ, 1983; Organ & Ryan, 1995), employee perceptions of fairness (Niehoff & Moorman, 1993), and personality traits (Organ & Ryan, 1995). In addition, researchers in the field have also examined the effects of OCB on employees and organizations. For example, the consequences include “managers’ reward allocation decisions” and “various indices of organizational performance” (Podsakoff et al., 2009). Thus, it is important to investigate OCB in order to improve organizational performance.
OCB is a spontaneous behavior which means that sanctions cannot easily enforce it (Smith, Organ, and Near, 1983). Therefore, what interests us are the factors can effectively influence OCB. Farh et al. (1990) propose that a perception of fairness can greatly affect OCB. Organizational justice is related to “perception of fairness” because it is a set of rules that states what is fair or unfair, and it is defined as the perception of fairness when
individuals react to fair or unfair treatment (Homans, 1961; Greenberg, 1990; Colquitt et al., 2001; Moorman, 1991; Colquitt et al., 2001). In addition, it has three dimensions: distributive, procedural, interpersonal and informational justice (Colquitt et al., 2001). Employees consider a firm as fair when it is able to distribute or allocate payoffs based on employees’ performance and outcomes. Organizational justice can impact job attitudes and job satisfaction, and further affect employees’ behavior, including organizational citizenship behavior (Greenberg, 1990; Greenberg & Tyler, 1987).
In the behavioral strategy field, individual differences such as individual work values, have been little explained when scholars investigate employees’ behavior within firms (Ros, Schwartz, & Surkiss, 1999). Previous studies have found that work values influence job satisfaction, performance and turnover intention (Ros et al., 1999; Feather & Rauter, 2004). In addition, these studies also state that work values can affect the expectations and understanding of employees on organizations. For example, employees’ willingness to be a valued employee might be enhanced if firms offer good salary and work conditions; that is to say, firms satisfy those employees who value a higher salary, and better work conditions could change their behavior or attitudes (Ros et al., 1999; Feather & Rauter, 2004).
However, there is not much research on how work values influence the behavior of employees, especially influence organizational citizenship behavior. The main question is: How do work values affect the decision-making process when people perform organizational citizenship behavior? Holtz and Harold (2013) state that work values are helpful to account for various behaviors related to the perception of fairness. Thus, the key research question of this study is to investigate the effects of work values on the relationship between justice and
organizational citizenship behavior.
This paper will make the following contributions. Firstly, on a theoretical level, it will fill the gap of identifying the effects of work values on the relationship between organizational justice and organizational citizenship behavior. Secondly, it also contributes on a practical level because managers can use these findings to improve a firm’s performance through enhanced OCB which is the result of organizational justice moderated by work values.
The paper is structured as follows. A critical literature review will be discussed in the first part. In the second part, research design will be outlined, followed by a results section. Subsequently, a discussion based on results will be introduced, including findings, implications, limitations and directions for future research. Finally, the main conclusions of the paper will be presented.
II Literature review
The following paragraphs are discussing the main insights of the existing literature on the topic. Firstly, I will introduce the nature of organizational citizenship behavior and the role of organizational justice. Then, I outline the main findings of the existing literature with respect to distributive, procedural justice and interactional justice and their influences on employees performing organizational citizenship behavior, and the mechanisms in explaining the effect of organizational justice on organizational citizenship behavior. Subsequently, I discuss the role of work values as an individual difference variable which affects human behavior, notably organizational citizenship behavior. Finally, literature gap and research question are outlined, followed by the conceptual model.
Organizational citizenship behavior (OCB) is an extra-role behavior that may not be
encouraged by organization’s formal reward system (Podsakoff, MacKenzie, Paine, & Bachrach, 2000). For example, OCB is the behavior that workers who complete their own jobs assist colleagues without being asked for help. OCB is used to evaluate the effects of employees’ work behaviors and defined as individual behavior and attitudes that will benefit organizations (Tansky, 1993; Greenberg, 1993; Smith, Organ, & Near, 1983; Bateman & Organ, 1983; Farh, Podsakoff, & Organ, 1990). Previous studies of OCB find that it has a positive influence on firm performance in that OCB promotes team coordination and team efficiency, and further assists firms to achieve goals (Posdakoff & Mackenzie, 1994).Thus, it is important to understand which factors contribute to OCBs within organizations.
Several researchers propose that variables such as job satisfaction and organizational identification have positive relations with organizational citizenship behaviors (Podsakoff et al., 2000). However, variables such as leader monitoring can restrict organizational citizenship behavior because supervisors focus more on in-role behavior rather than extra-role behavior due to task accomplishment (Niehoff & Moorman, 1993). That is to say, managers encourage employees to perform OCB under the precondition of accomplishment of their own work. Thus, deep insight in the nature of organizational citizenship behavior is needed here to understand how to positively facilitate organizational citizenship behavior and thereby enhance organizational performance. Organ (1990) proposes that perception of fairness can greatly affect OCB and it is a good predictor for employees displaying organizational citizenship behavior. The subsequent paragraphs will discuss organizational justice and its influences on organizational citizenship behavior.
Organizational justice is the individual perception of fairness within organizations
(Greenberg, 1987). According to Greenberg (1990), organizational justice is able to ensure organizational efficiency and employee satisfaction. In the organizational justice literature, researches find that organizational justice is relevant to organizational behavior such as job performance, job satisfaction, organizational commitment, trust and organizational citizenship behavior; that is to say, employees feel happier and achieve higher performance when they perceive a treatment as fair (Colquitt, Conlon, Wesson, Porter, & Ng, 2001; Colquitt & Greenberg, 2003; Zapata-Phelan et al., 2009). Researchers propose that organizational justice is positively related to organizational citizenship behavior and that employees who perceive higher justice will conscientiously work and help other colleagues (Skarlicki & Latham, 1997; Tansky, 1993; Greenberg, 1993). This leads to our first hypothesis.
Hypothesis 1: Organizational justice has a significant positive influence on OCB.
Organizational commitment is an important factor to realize the mechanism behind organizational justice toward OCB (Meyer, Allen & Smith, 1993; Folger & Konovsky, 1989). Moreover, Steers and Porter (1979) define organizational commitment as attitudes or behavioral intension with respect to long-term perspective on membership within an organization, involving the sense of identification of the values and goals of organizations, the willingness to effort to the interest of organizations, and the strong desire to stay members of organizations (Porter, Steers, Mowday & Boulian, 1974). Meyer and Allen (1991) specify three forms of organizational commitment, which are affective, continuance, and normative commitment. These types of commitment will impact choice and behavior toward remain in
the organization. Affective commitment is relevant to the relationship between work experience and satisfaction of needs; while normative commitment focuses on the sense of loyalty of employees to employers or the sense of obligation to reciprocate (Meyer, Allen & Smith, 1993). Finally, continuance commitment depends on comparable options outside the organization; that is to say, continuance commitment cannot be affected by things within an organization, i.e. organizational justice. With regard to organizational justice, it is significantly associated with organizational commitment and it is a predictor of organizational commitment (Folger & Konovsky, 1989; Lowe & Vodanovich, 1995). Furthermore, organizational commitment is an antecedent which can predict the extent to which employees display OCB (O’Reilly and Chatman, 1986; Williams and Anderson, 1991).That is to say; Organizational justice might promote the emergence of organizational citizenship behavior through organizational commitment. As suggested by Meyer and Allen (1991), normative and affective commitment could positively affect organizational citizenship behavior; therefore, this paper will focus on affective and normative commitment. This leads to following hypotheses:
Hypothesis 2a: Affective commitment is positively associated with OCB. Hypothesis 2b: Normative commitment is positively associated with OCB.
Hypothesis 3a: Affective commitment has mediating effect on the relationship between organizational justice and OCB.
Hypothesis 3b: Normative commitment has mediating effect on the relationship between organizational justice and OCB.
procedural justice and interactional justice and it is an open question whether each dimension positively influence organizational citizenship behavior (Adams, 1965; Homans, 1961; Thibaut & Walker, 1975; Leventhal, 1976; Bies & Moag, 1986; Colquitt et al., 2001). We therefore discuss the differences among distributive, procedural and interactional justice, and their influences on organizational citizenship behavior, respectively.
Distributive justice
Distributive justice is defined as the “fairness of outcome distributions or allocations”
that workers perceive (Homans, 1961; Colquitt et al., 2001). In other words, fairness is the perception of employees that determine employee reactions to outcomes, which are compared with their colleagues. However, employees not only pay attention to their payoffs, but also calculate the ratio of contributions to outcomes and compare the ratio with other people (Adams, 1965). This is called the equity theory. Furthermore, Leventhal (1976) proposes that, apart from equity rule, other rules such as equality and need are also influence people’s perception of fairness, which is called justice judgment model (cited by Coliquitt et al., 2001; Greenberg, 1990).
Previous research of organizational justice has stated that distributive justice influences OCB (Farh et al., 1990; Niehoff & Moorman, 1993). In other words, employees who evaluate the distributive outcomes as fair will shape their behavior to benefit the organization even beyond their job duties (Niehoff & Moorman, 1993).
Therefore, this leads to the following hypothesis.
Hypothesis 4a: distributive justice will positively influence organizational citizenship behavior
Procedural justice
Another kind of justice is classified as procedural justice and it focus on the process of distribution decision-making (Thibaut & Walker, 1975; Leventhal, 1976). Thibaut and Walker (1975) find that the perception of fairness is derived from not only outcomes of distribution, but also the process of establishing the process of distributing. Specifically, their study finds that, if people who decide the distribution result are subjective, workers will evaluate the distribution or allocation as unfair. In other words, employees question the participants in distributing process and they doubt the fairness of the process. In contrast, people consider the procedure as fair when they can control the process, and this is called process control (Thibaut & Walker, 1975). For example, reward systems need to be open in the sense that employees have right to participate in the process and allow them to frame the system, and transparent, in the sense that people know how it works (Steel & MacDonnell, 2012). In this example, people have a perception of procedural fairness through process control and gain a sense of autonomy in building reward system. Furthermore, there are six rules of procedural justice developed by Leventhal, which is called Leventhal criteria, consisting of the consistency rule ( consistency over time and consistency between persons), the representativeness rule (represent the interests from different groups in an organization), the bias suppression rule (avoid intervention of personal interests in decision-making), the accuracy rule (the information used to decision-making is accurate), the correctability rule (the decision-making process can be modified ), and the ethicality rule (conform to standards of social norms and ethics) (Skarlicki & Folger, 1997; Colquitt et al., 2001; Niehoff & Moorman, 1993). These rules are capable of helping organizations to maintain fairness of
process.
Procedural justice can create a harmonious work place that promote employees to perform organizational citizenship behavior (i.e. help other colleagues) (Lambert & Hogan, 2013). In other words, fair procedures send a signal to employees that justice is valued for a organization, and create a sense of cooperation among colleagues, leading to higher involvement in extra-role behaviors, and in particular, organizational citizenship behavior.
Therefore, this leads to the following hypothesis
Hypothesis 4b: Procedural justice will positively influence organizational citizenship behavior.
Interactional justice
The last type of organizational justice is interactional justice that reflects how personal interaction influences organizational justice in the process of distribution and execution (Bies & Moag 1986). It focuses on leader-member relation, that is, how employees react to their supervisors (Colquitt et al., 2001). Scholars argue that interactional justice is a social form of procedural justice (Cropanzano & Greenberg, 1997). In other words, procedural justice focuses on the process of distribution or allocation, while interactional justice focuses on the communication process and especially concern on the behavior of supervisors (Moorman, 1991). In addition, employees tend to feel respect and dignity from their supervisors when they are under higher interactional justice, leading to a higher level of organizational identification of employees (Zhao, Peng, & Chen, 2014). Organizational identification is “the degree to which a person defines him or herself as having the same attributes that he or she believes define the organization” (Dutton, Dukerich, & Harquail, 1994, p. 239). That is to say,
interactional justice can influence OCB through organizational identification. Moreover, organizational identification is a form of organizational commitment.
Therefore, this leads to the following hypothesis.
Hypothesis 4c: Interactional justice will positively influence organizational citizenship behavior.
2. Work values, organizational justice and organizational citizenship behavior
Work values are personal values toward work that can influence attitudes and behavior (Ros, Schwartz & Surkiss, 1999; Kabasakal, Dastmalchian & Imer, 2011). Elizur, Borg, Hunt and Beck (1991) identify three dimension of work values based on “modality of outcome” into intrinsic (or cognitive group), extrinsic (or instrumental group or material work values) and social (or affective group) work values. The “modality of outcome” indicates the outcomes that employees require toward work. Overall, work values are subjective attitudes that reflect individual preferences and principles.
Work values have been proved to be predictive of organizational citizenship behavior in work settings (Feather & Rauter, 2004; Kabasakal et al., 2011; Chen & Kao, 2012). In other words, employees perform OCB to extend opportunities that satisfy their own work values. For example, people might engage in OCB to reach their goals relating to specific values, depending on the importance of the work values (Feather & Rauter, 2004). Consequently, work values are important to investigate because they reflect the intention of employees to perform OCBs (Feather & Rauter, 2004).
To understand the mechanism behind work values on affecting OCBs, we discuss the differences among intrinsic, extrinsic, social and prestige work values and their influences on
organizational citizenship behavior.
With respect to the motivation theory, work values are regarded as a particular kind of motives that motivate people to implement behavior toward specific goals (Elizur et al., 1991; Kabasakal et al., 2011). Specifically, people shape their behavior based on different work values, including intrinsic, extrinsic, social and prestige work values. For example, Kabasakal et al., (2011) propose that employees who attach importance to social work values are likely to perform “prosocial behavior”, namely, organizational citizenship behavior.
Furthermore, based on needs theory formulated by Alderfer (1972), which is developed from Maslow’s hierarchy of needs, behavior is driven by needs that needs satisfaction can motivate people to perform behavior (Elizur et al., 1991). To be specific, existence (E), relatedness (R) and growth (G) are corresponding to extrinsic, social and prestige, and intrinsic work values, respectively. In Alderfer’s ERG theory, existence needs cannot motivate employees to perform well while relatedness and social needs are positively associated with job motivation (Schmitt & White, 1978). That is to say, intrinsic, social, and prestige work values can encourage employees to engage in OCB. As two-factor theory by Herzberg (1974) stated, motivators can encourage people while hygiene cannot motivate people. In this theory, motivators reflect the intrinsic (cognitive) work values whereas hygiene represents the extrinsic (material) and social work values (social items) (Elizur et al., 1991). Additionally, there are no arguments based on two-factor theory that prestige work values have influence on OCB. In contrast to the ERG theory mentioned above, social work values are somehow different that social work values are regarded as hygiene, which cannot motivate people. In other words, intrinsic work values can be identified as a motivation
variable that encourages employees to perform OCB whereas extrinsic work values and social work values cannot enhance OCB. To investigate whether social work values have effect on OCB, social work values are served as a factor that can positively associated with OCB. In conclusion, intrinsic, social and prestige work values can positively affected OCB while extrinsic work values cannot affect OCB.
Therefore, this leads to following hypotheses.
Hypothesis 5a: Employees that find that intrinsic work values are more important will engage in more OCB.
Hypothesis 5b: Employees that find that social work values are more important will engage in more OCB.
Hypothesis 5c: Employees that find that prestige work values are more important will engage in more OCB.
With respect to organizational justice, Holtz and Harold (2013) state that work values are important in the organizational justice literature because employees’ values and beliefs might influence behavioral action based on a sense of justice. Furthermore, some researchers propose that work values are positively related to OCB (Ryan, 2002; Feather & Rauter, 2004); while sense of justice can positively influence OCB. However, extrinsic work values do not have influence on the extent to which employees engage in OCB (Schmitt & White, 1978). Therefore, this leads to the hypotheses.
Hypothesis 6a: The positive effect of organizational justice on OCB becomes more positive when the employees find that intrinsic work values are more important than when the employees find that intrinsic work values are less important.
Hypothesis 6b: The positive effect of organizational justice on OCB becomes more positive when the employees find that social work values are more important than when the employees find that that social work values are less important.
Hypothesis 6c: The positive effect of organizational justice on OCB becomes more positive when the employees find that prestige work values are more important than when the employees find that prestige work values are less important.
3. Literature gap and research question
Organizational justice can efficiently and effectively foster organizational citizenship behavior (Greenberg, 1990; Greenberg & Tyler, 1987). However, existing literature neglects the influence of work values on behavior of employees because there are many types of values and there is no uniform standards to describe work values (Holtz & Harold, 2013). Therefore, this paper would like to discuss the effects of work values on the relationship between justice and organizational citizenship behavior.
The purpose of this thesis is to understand the mechanism of organizational citizenship behavior under different dimensions of organizational justice, while taking into account the moderating effect of work values. Therefore, the research question is as follow:
What is the moderating effect of work values on the relationship between organizational justice (distributive, procedural and interactional) and organizational citizenship behavior (OCB)?
This research will make several contributions. On a theoretical level, it fills the gap of identifying the effects of work values on the relationship between three dimensions of organizational justice and organizational citizenship behavior. In addition, it also contributes
on a practical level because managers realize what kind of justice they should strengthen to foster organizational citizenship behavior. Moreover, managers can use these findings to improve firm performance through enhanced OCB which is the result of organizational justice, moderated by work values.
The conceptual model is shown in Fig 1. We argue that organizational justice can positively affect OCB (H1). In addition, work values can enhance the relationship between organizational justice and OCB.
Figure 1 – Conceptual Model: Organizational justice, work values, organizational
commitment and OCB
Organizational Justice: - Distributive Justice - Procedural Justice - Interactional Justice
Organizational Citizenship Behavior: Extra-role behavior that benefit to organization Organizational Commitment: - affective commitment - normative commitment Work Values: - Intrinsic - Extrinsic - Social - Prestige Organizational Justice: - Distributive Justice - Procedural Justice - Interactional Justice
Organizational Citizenship Behavior: Extra-role behavior that benefit to organization Control Variables: - Age - Gender - Level of Education - Organizational tenure
III Methodology
The following paragraphs will explain the research design of this study. First, the research technique – a traditional survey questionnaire - is outlined, including the description of the sample and the data collection procedure. Subsequently, a description of the measurement of the variables mentioned in the previous sections is introduced.
1. Sample
A survey that included measurement of work values, organizational justice, organizational commitment and organizational citizenship behavior was mailed to 528 bank employees in Hulun Buir, PR. China. Each participant answered the survey through Qualtrics.com. The two banks were chosen because they were almost similar in most relevant ways (in terms of size, industry, number of employees and global location). One is Bank of China Hulunbuir Branch, which had 278 employees; while another is Industrial and Commercial Bank of China Hulunbuir Branch, which had 486 employees. Manager of the former bank sent email to all employees, while manager the latter bank sent emails to the first 250 employees of the list of staff based on the last name.
The survey was filled out by 261 respondents and with available data by 250 respondents, resulting in a response rate of 49.4%. Among the data, most were female (60.8 %) with an average age of 36 years and they averaged nearly 5 to 10 years of experience working in the bank. A majority had completed college (99.6%).
2. Procedure
The survey was prepared in Chinese. The original questionnaires from existing literature were in English. I translated the questionnaires into Chinese and checked the back-translation
from another proficient English speaker to ensure that the Chinese version best reflects the original English version. Finally, I also modified the last version of Chinese Questionnaires.
In addition, pilot testing was used to ensure the reliability and validity of the items in the Chinese questionnaire. Chinese friends and relatives were required to fill the questionnaire since they represented the mainstream of employees, ranging from 20 to 45.
At the beginning of the survey, there was an invitation letter to gain permission to collect data and use the data in my thesis. In addition, it was mentioned that all the data would be kept confidential and were only used for academic purpose.
Participants first completed the questionnaire of work values, then of organizational justice and then of organizational commitment, subsequently of OCB, and finally they were asked to provide demographic data.
3. Measures
3.1. Independent Variables
The scale from Niehoff and Moorman (1993) was used to investigate organizational justice, including three dimensions measuring perceptions of distributive justice, of procedural justice and of interactional justice. The questionnaire contains 20 items; respondents are required to rate each item on a scale from 1 (strongly disagree) to 7 (strongly agree). Distributive justice (5 items; α=.872) measures fairness of work outcomes.
Procedural justice (6 items; α=.864) measures the degree to which employees’ job decisions
are made through information transfer and employees’ feedback. Interactional justice (9 items; α=.950) measures employees’ feeling towards the job decision making process. The items of the questionnaire measuring organizational justice are summarized in Table 1 below.
Table 1 – Items for organizational justice
Item Type
1. My work schedule is fair. Distributive 2. I think that my level of pay is fair. Distributive 3. I consider my work load to be quite fair. Distributive 4. Overall, the rewards I receive here are quite fair Distributive 5. I feel that my job responsibilities are fair. Distributive 6. Job decisions are made by the general manager in an unbiased
manner. Procedural
7. My general manager makes sure that all employee concerns are
heard before job decisions are made Procedural 8. To make job decisions, my general manager collects accurate
and complete information. Procedural
9. My general manager clarifies decisions and provides additional
information when requested by employees. Procedural 10. All job decisions are applied consistently across all affected
employees. Procedural
11. Employees are allowed to challenge or appeal job decisions
made by the general manager. Procedural 12. When decisions are made about my job, the general manager
treats me with kindness and consideration. Interactional 13. When decisions are made about my job, the general manager
treats me with respect and dignity. Interactional 14. When decisions are made about my job, the general manager
is sensitive to my personal needs. Interactional 15. When decisions are made about my job, the general manager
deals with me in a truthful manner. Interactional 16. When decisions are made about my job, the general manager
shows concern for my rights as an employee. Interactional 17. Concerning decisions made about my job, the general
manager discusses the implications of the decisions with me. Interactional 18. The general manager offers adequate justification for
19. When making decisions about my job, the general manager
offers explanations that make sense to me. Interactional 20. My general manager explains very clearly any decision made
about my job. Interactional
3.2. Dependent Variable
The 7-point scale of Lee and Allen (2002) was used to measure OCB. This questionnaire consists of 16 items and respondents are required to respond “how often you engage in these behaviors” on a scale from 1 (never) to 7(always). It contains 16 items (α=.889) and each item represents one behavior in the work setting, which is beneficial to both the individual and the organization. The items of this scale are outlined in Table 2.
Table 2 – Items for OCB Item
1. Helps others who have been absent.
2. Willingly give your time to help others who have work-related problems.
3. Adjust your work schedule to accommodate other employees’ requests for time off. 4. Go out of the way to make newer employees feel welcome in the work group.
5. Show genuine concern and courtesy toward coworkers, even under the most trying business or personal situations.
6. Give up time to help others who have work or non work problems. 7. Assist others with their duties.
8. Share personal property with others to help their work. 9. Attends functions not required that help company image. 10. Keep up with development in the organization.
11. Defend the organization when other employees criticize it. 12. Show pride when representing the organization in public.
13. Offer ideas to improve the functioning of the organization. 14. Express loyalty toward the organization.
15. Take action to protect the organization from potential problems. 16. Demonstrate concern about the image of the organization.
3.3. Mediating Variable
To measure organizational commitment, scales developed by Meyer, Allen and Smith (1997) were used. For the purpose of this research, 12 items of their questionnaire (36 items in total) were chosen as explained below. Meyer et al. (1997) specify 6 factors and each factor contains 6 items. The factor 1, 2 and 3 reflect affective, continuance, and normative commitment to the specific nurse occupation, and the factor 5 reflects continuance commitment, which can be affected by options outside the company and cannot be influenced by organizational justice. Hence, items from factor 4 and factor 6 were chosen as the measurement. Factor 4, which is a subscale of affective commitment (6 items, α=.831) measures the affective attachment to the organization through the consistency of experience and expectation within the organization and the satisfaction of needs. Factor 6, which is a subscale of normative commitment (6 items, α=.809) measures the attachment to one’s employers and the organization through a sense of obligation to stay. Each respondent responded the question “how you feel about your company” from 1 (strongly disagree) to 7 (strongly agree).
Table 3 – Items for affective commitment and normative commitment
Item Type
1. I would be very happy to spend the rest of my career with this organization.
2. I really feel as if this organization’s problems are my own. Affective 3. I do not feel a strong sense of “belonging” to my organization. Affective 4. I do not feel “emotionally attached” to this organization. Affective 5. I do not feel like “part of the family” at my organization. Affective 6. This organization has a great deal of personal meaning for me. Affective 7. I do not feel any obligation to remain with my current employer. Normative 8. Even if it were to my advantage, I do not feel it would be right
to leave my organization now.
Normative
9. I would feel guilty if I left my organization now. Normative 10. This organization deserves my loyalty. Normative 11. I would not leave my organization right now because I have a
sense of obligation to the people in it.
Normative
12. I owe a great deal to my organization. Normative
3.4. Moderating Variable
The scale used to measure work values, a survey from Ros, Schwartz and Surkiss (1990), is adapted in this study. The original survey asked respondents to rate the 10 items on a scale from 1 (very important) to 4(not at all important) to respond to the question “how important is each of the following to you in choosing an occupation?”. However, other three questionnaires are on 7-point likert scale while questionnaire of work values is on 4-point likert scale. To modify all the questionnaires on the same likert scale, 7-point Likert scale was used to measure work values from 1 (very important) to 7 (not at all important). In addition, 7-point likert scale provides a wider range options for participant to choose from so that opinions of participant are better evaluated. The first item (good salary and work conditions) was separated into two items (item 1 is good salary and item 2 it work conditions) because
salary and work conditions are separate items for Chinese employees. Its reason lies in the different requirement of applicants. For most job candidates, they could bear bad work conditions to earn fat salary. However, for some other job candidates, they may look for a low-paid job, which has a quality and harmonious workplace. The resulting 11 items are selected to represent the four types of work value. Social work values are measured with three items (α=.593): “contributing to people and society”, “work with people”, and “social contact with co-workers”. Extrinsic work values are measured with 3 items (α=.797): “good salary”, “work conditions” and “job security”. Prestige work values are assessed with two items (α=.195): “authority to make decisions over people” and “prestigious, highly valued work”. Intrinsic work values are assessed with two items (α=.398): “interesting and varied work” and “work in which you are your own boss”. The last item is “opportunities for occupational advancement”, which measures mixed work values. The values of Cronhnbach’s Alpha for intrinsic work values, social work values, and prestige work values were less than .70; although the values are low, these variables are still employed since they has been used like this in early research, and to investigate whether low reliability can influence the results. The explanation of this poor reliability is presented in the Reliability Analysis in the Results part.
Table 4 – Items for work values
Item Type
1. Good salary Extrinsic
2. Work conditions Extrinsic 3. Job security (permanent job, pension) Extrinsic
4. Interesting and varied work Intrinsic 5. Work with people Social 6. Prestigious, highly valued work Prestige 7. Work in which you are your own boss Intrinsic 8. Contributing to people and society Social 9. Authority to make decisions over people Prestige 10. Social contact with co-workers Social 11. Opportunities for occupational advancement Mixed
3.5. Control Variables
This research controls for respondents’ personal profile, i.e., age (1=female, 2=male), gender, level of education, and organizational tenure in the organization as the existing studies outline.
In this paper, age was measured by a multiple choice question which contained answers from age 18 to 66 that respondents can choose from the dropdown list.
As Tremblay, Sire and Balkin (2000) mentioned in their article, they regard the level of education as an individual difference variable that has been shown to influence perceptions. Hence it has been included as a control variable. When comparing the degrees in China and Western countries, there are slight differences between the Western and Chinese Bachelor’s degree and college level. Chinese college students take almost three years to finish their study and they focus on practice and specialized skills, whereas Chinese bachelor students take four academic years to complete their study and they focus on theoretical research.
With respect to organizational tenure, other studies propose that job attitudes are related to tenure while organizational citizenship behavior is defined as an individual attitude toward
work (Herman & Hulin, 1972; Pfeffer 1980). Hence, organizational tenure was considered as a control variable.
The table of control variables is outlined below.
Table 5– Control Variables
NO. Control variable Items
1. Age (years) - Less than 18 - 18 - 19 - 20 - … - 65 - 66 2. Gender - Female - Male
3. Level of education - Senior High school level - College level
- Bachelor’s level - Master’s level - Doctorate level - others
4. Organizational tenure - less than 3 months - between 3 and 6 months - between 6 and 12 months - between 1 and 2 years - between 2 and 3 years - between 3 and 4 years - between 5 and 10 years - more than 10 years
IV Results
The following paragraphs present the results of this research. Firstly, data are cleaned for the purpose of this study. Then, an overview of the data is presented. Subsequently, the results of the reliability analysis for all variables are reported. Next, a correlation analysis is performed to report the most significant correlations. Subsequently, the normality analysis is reported. Finally, the results of the regressions analysis are outlined in order to test the
hypotheses which were formulated in the literature review.
1. Data Cleaning
For the purpose of this research, data were cleaned at the first stage. Entries with values that were all 7 or 1 were deleted since these data are useless because participants might not have paid attention to the survey. Moreover, after running the frequencies command, the output showed that there were only three missing values in age, which is regarded as a control variable. Hot deck imputation could not be used to replace missing data since age is not correlated to other variables, which means it is not possible to find a deck variable (Myers, 2011). Thus, the three entries with missing ages were deleted. Then, the frequencies command was rerun to ensure all missing values had been replaced.
After cleaning the data, the data, the number of valid data was 250.
2. Descriptive Statistics
In this study, three categorical variables were used, namely the control variables gender,
level of education and organizational tenure.
The description of categorical variables is presented in Table 6 – Descriptive Statistics
for Categorical Variables.
Concerning the gender, the survey was completed by 152 (60.8%) females and 98 (39.2%) males.
With regard to the level of education, only 1 (0.4%) had a Senior High School level, while 50 (20.0%) had a College level. The majority of the respondents, 168 (67.2%), had obtained a Bachelor’s level. Twenty-nine (11.6%) respondents had a Master’s level. There were 2 (0.8%) respondents with a Doctorate level,.
Among the 250 respondents, 3 (1.2%) people had worked in the organization for no more than 3 months and 2 (0.8%) had worked there between 3 and 6 months. Eight (3.2%) had between 6 and 12 months of work experience, 11 (4.4%) people had between 1 and 2 years of work experience, while 15 (6.0%) had between 2 and 3 years of work experience. There were 46 (18.4%) people who had worked between 3 and 4 years for the organization, and 93 (37.2%) respondents from the sample had worked there between 5 and 10 years, which is the first modus. Finally, 72 (28.8%) people had been employed in their organization for more than 10 years.
Table 6 – Descriptive Statistics for Categorical Variables
Variable Level Frequency % Valid % Cumulative %
Gender - Female - Male 152 98 60.8 39.2 60.8 39.2 60.8 100.0 Level of Education - Senior High School level
- College level - Bachelor’s level - Master’s level - Doctorate level - others 1 50 168 29 2 0 0.4 20.0 67.2 11.6 0.8 0.0 0.4 20.0 67.2 11.6 0.8 0.0 0.4 20.4 87.6 99.2 100.0 0.0 Organizational Tenure
- less than 3 months - between 3 and 6 months - between 6 and 12 months - between 1 and 2 years - between 2 and 3 years - between 3 and 4 years - between 5 and 10 years - more than 10 years
3 2 8 11 15 46 93 72 1.2 0.8 3.2 4.4 6.0 18.4 37.2 28.8 1.2 0.8 3.2 4.4 6.0 18.4 37.2 28.8 1.2 2.0 5.2 9.6 15.6 34.0 71.2 100.0
Apart from the categorical variables, eleven continuous variables were used, namely the independent variables – distributive justice, procedural justice and interactional justice, the dependent variable – organizational citizenship behavior, the moderating variables – intrinsic,
extrinsic, social and prestige work values, the mediating variables – affective commitment and normative commitment, and the control variable – age. The description of continuous
variables is presented in Table 7 at the end of this section.
With respect to the independent variable, distributive justice had values between 1.60 and
7.00, with a mean of 5.15 and a standard deviation of 0.94. The scores of procedural justice
were ranging from 2.50 to 7.00, with a mean of 4.84 and a standard deviation of 0.90. For
interactional justice, the values were between 1.78 and 7.00, with a mean of 5.20 and a
standard deviation of 0.95
In terms of organizational citizenship behavior, the values of the dependent variable were between 3.56 and 7.00, with a mean of 5.10 and a standard deviation of 0.72
The results of the moderating variable intrinsic work values had values ranging from
1.00 to 7.00, with a mean of 4.51 and a standard deviation of 1.17 (Table 2). The values for extrinsic work values raged from 1.00 to 7.00, with a mean of 6.26 and a standard deviation
of 0.84.When the moderating variable is social work values, the values were between 1.00 and 7.00, with a mean of 4.70 and a standard deviation of 1.09. For the moderation variable,
prestige work values, the scores were between 1.00 and 7.00, with a mean of 4.48 and a
standard deviation of 0.97 (Table 2).
With regard to the control variable age, the values were ranging from 21 to 52, with a mean of 36.51 and a standard deviation of 7.09.
Table 7 - Descriptive Statistics for Continuous Variables
Variable N Min. Max. Mean. Standard Deviation
Distributive Justice 250 1.60 7.00 5.15 0.94 Procedural Justice 250 2.50 7.00 4.84 0.90
Interactional Justice 250 1.78 7.00 5.20 0.95
OCB 250 3.56 7.00 5.10 0.72
Intrinsic Work Values 250 1.00 7.00 4.51 1.17 Extrinsic Work Values 250 1.00 7.00 6.26 0.84 Social Work Values 250 1.00 7.00 4.70 1.09 Prestige Work Values 250 1.00 7.00 4.48 0.97 Affective Commitment 250 2.00 7.00 5.29 1.06 Normative Commitment 250 1.83 7.00 5.34 0.89
Age 250 21 52 36.51 7.09
3. Correlation Analysis and Reliability Analysis
The correlations among study variables are outlined in Table 8– Correlations and
Reliabilities at the end of this section. The significant correlations at the 0.05 level have been
marked with an asterisk (*) and the significant correlations at the 0.01 level have been marked with two asterisks (**). These correlations were calculated using the Pearson product-moment correlation coefficient.
The table of correlations contains the following set of variables: intrinsic work values,
extrinsic work values, social work values, prestige work values, distributive justice, procedural justice, interactional justice, affective commitment, normative commitment, organizational justice, age, gender, level of education, and organizational tenure.
In the following paragraphs, the most important correlations with significant coefficients at or below the 0.05 level are discussed. Firstly, correlations between the moderating variables and dependent variable are outlined. Subsequently, the correlations between the dependent variable and mediating variables are discussed. Then, the most important correlations between dependent variable and control variables are represented. Subsequently, some of the correlations between control variables and moderating variables are reported. Finally, the correlations between control variables and mediating variables are outlined.
As the table shown, the moderating intrinsic work values was positively correlated with the dependent variable OCB (Pearson correlation coefficient =0.393; N = 250; Sig. =0.000). Similarly, moderating social work values had a positive correlation with OCB (Pearson
correlation coefficient = 0.423; N = 250; Sig. =0.000); while moderating prestige work values also had positive relation with OCB (Pearson correlation coefficient = 0.282; N = 250; Sig. =0.000). However, the moderating work values was not correlated with OCB (Pearson correlation coefficient = 0.096; N = 250; Sig. =0.132). These findings correspond with what
hypothesis 5a, 5b, and 5c hypothesized.
With respect to the mediating variable, affective commitment, OCB was positively correlated with it (Pearson correlation coefficient = 0.199; N = 250; Sig. =0.002); while another mediating variable, normative commitment, was also correlated positively with the dependent variable OCB (Pearson correlation coefficient = 0.273; N = 250; Sig. =0.000).
Furthermore, the analysis presented positive correlations between the dependent variable
OCB and control variables. Level of education was positively correlated with OCB (Pearson correlation coefficient = 0.347; N = 250; Sig. =0.000). Similarly, there was a positive
correlation between organizational tenure and OCB (Pearson correlation coefficient = 0.155;
N = 250; Sig. =0.014). The significant results indicated that educational level and work
experience could affect OCB.
Additionally, regarding the correlations between the control variables and the moderating variables, organizational tenure was negatively correlated with moderating intrinsic work
values (Pearson correlation coefficient = -0.152; N = 250; Sig. =0.016), while organizational tenure did not correlate with other moderating variables.
In addition, the correlations in Table 8 showed that level of education was positively with mediating affective commitment (Pearson correlation coefficient = 0.373; N = 250; Sig.
=0.000) and mediating normative commitment (Pearson correlation coefficient = 0.345; N = 250; Sig. =0.00).
For the purpose of this study, ten variables were measured using multiple questions: the dependent variable – organizational citizenship behavior, the independent variable –
organizational justice, the moderating variable – work values, and the mediating variable – organizational commitment. The Cronbach’s alpha coefficients of all variables are outlined in Table 8 at the end of this section.
The dependent variable organizational citizenship behavior was captured using 16 questions and the value of the Cronbach’s alpha was 0.889 (Table 8).
The independent variable organizational justice was captured in three dimensions; distributive justice using 5 items, procedural justice using 6 items, and interactional justice using 9 items. The Cronbach’s alpha coefficient of distributive justice was 0.872, the Cronbach’s alpha coefficient of procedural justice was 0.864, and the Cronbach’s alpha coefficient of interactional justice was 0.950, which are all considered to be a good reliability.
Concerning the moderating variables, the Cronbach’s alpha coefficient for intrinsic work
values (2 items) was α = 0.398, the Cronbach’s alpha coefficient for extrinsic work values (3
items) was α =0.797, the Cronbach’s alpha coefficient for social work values (3 items) was α = 0.593, and the Cronbach’s alpha coefficient for prestige work values (2 items) was α = 0.195. The Cronbach’s alpha coefficients for intrinsic work values, social work values, and
prestige work values were less than 0.70. To explain these exceptional cases, three possible reasons can be introduced. Firstly, items used to measure the preference of Chinese employees when choosing jobs were not appropriate since these items were formulated based on Western people. Another reason is the small number of items that 11 items were used to measure four types of work values, that is, each dimension of work values contained only two to three items. However, even the reliabilities of intrinsic work values, social work values and prestige work values are poor; these variables are still employed since the previous study also used this measurement.
In the case of the mediating variables – affective commitment and normative commitment, the scores of the Cronbach’s alpha were 0.831 and 0.809, respectively.
Table 8–Correlations and Reliabilities
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14
1. Intrinsic work values (.398)
2. Extrinsic work values .258** (.797)
3. Social work values .642** .171** (.593)
4. Prestige work values .532** .170** .577** (.195)
5. Distributive justice .182** .114 .281** .283** (.872)
6. Procedural justice .278** .074 .387** .308** .671** (.864)
7. Interactional justice .194** .210** .324** .291** .653** .757** (.950)
8. Affective commitment .045 .428** .136* .034 .335** .281** .505** (.831)
9. Normative commitment .030 .290** .185** .179** .546** .510** .679** .764** (.809)
10. Organizational citizenship behavior .393** .096 .423** .282** .258** .431** .371** .199** .273** (.889)
11. Age -.253** .082 -.260** -.167** -.210** -.170** -.181** -.026 -.069 -.035 (-)
12. Gender -.066 .060 .001 -.040 -.024 .013 .015 .123 .106 -.022 .092 (-)
13. Level of education .278** .193** .279** .146* .255** .286** .343** .373** .345** .347** -.301** .088 (-)
14. Organizational tenure -.152* .054 -.099 -.021 -.075 .036 .044 .091 .058 .155* .704* .102 -.172** (-) ** Correlation is significant at the 0.01 level (2-tailed).
4. Normality Analysis
The normality analysis was carried out to test whether the variables OCB, distributive
justice, procedural justice, interactional justice, intrinsic work values, extrinsic work values, social work values, prestige work values, affective commitment and normative commitment
were normally distributed. The results of normality analysis are summarized in Appendix B,
Table 10.
In the case of the independent variable, distributive justice, had a mean value of 5.15 and a 5% trimmed mean of 5.18 (Table 9). This indicated that the more extreme scores did not have a significant influence on the mean. Given the negative skewness (-0.718) and positive kurtosis (0.378), the scores were clustered to the right and their distribution was peaked in the centre of the graph with a tail. Procedural justice had a mean value of 4.84 and a 5% trimmed mean value of 4.84 (Table 9), meaning that the more extreme scores did not have a significant impact on the mean. According to the negative skewness and kurtosis, the scores were clustered to the right and the distribution was relatively flat with many cases in the extremes. The independent variable interactional justice had a mean of 5.20 and a 5% trimmed mean of 5.23 (Table 9). This pointed out that extreme values did not affect the mean. The skewness values of interactional justice was negative (-0.682) and the kurtosis was positive (0.863), which indicated that the scores of the variable were clustered to the right and their distribution was peaked, clustered in the centre (Appendix B, Graph 1,2 and 3).
With respect to dependent variable, OCB, had a mean value of 5.10 and a 5% trimmed mean of 5.08, indicating that the more extreme scores did not have a significant impact on the mean (Table 9). The positive skewness (0.292) and the slightly negative kurtosis (-0.019) of
OCB indicated that the scores of the dependent variable were clustered to the left at the high-end values and their distribution was relatively flat with too many cases in the extremes (Appendix B, Graph 4).
The mediating variable affective commitment had a mean value of 5.29 and a 5% trimmed mean of 5.32 (Table 9). This showed that the more extreme scores did not have a substantial influence on the mean. Affective commitment had a negative skewness (-0.397) and a kurtosis (-0.566), meaning that the scores were clustered to the right and their distribution was peaked, clustered in the centre (Appendix, Graph 9). With regard to
normative commitment, the mediating variable had a mean value of 5.34 and a 5% trimmed
mean of 5.37 (Table 9). Furthermore, the negative skewness (-0.632) and positive kurtosis (0.483) indicated that the scores were always clustered to the right at the high-end values and their distribution was peaked, clustered in the centre (Appendix, Graph 10).
The moderating variable work values, had mean values ranging from 4.48 to 6.26 and 5% trimmed means ranging from 4.51 to 6.37 (Table 9). This showed that the more extreme scores did not have a significant influence on the mean. Intrinsic work values had a negative skewness (-0.275) and a positive kurtosis (0.142), meaning that the scores were clustered at the right-hand side of the graph and their distribution was peaked in the centre. Similar to intrinsic work values, extrinsic work values also had negative skewness (-2.523) and positive kurtosis (9.431), indicating that the scores were clustered to the right and their distribution was peaked, clustered in the centre of the graph with a thin tail. In the case of social work
values, negative skewness and kurtosis pointed out that the scores were clustered to the right
of prestige work values were clustered to the right at the high end side of the graph and their distribution was peaked, clustered in the centre (Appendix B, Graph 5, 6, 7 and 8).
The results of the Kolgomorov-Smirnov statistics are outlined in Appendix B, Table 10. The test of Kolgomorov-Smirnov assessed the normality of the distribution of scores. The values for three independent variables were significant, which indicated a violation of the assumption of normality (Appendix B, Table 10). However, the normal probability plots showed that there was a reasonably straight line in all plots of the three variables, indicating the relatively normal distribution of the scores (Appendix B, Graph 11, 12 and 13).
In the case of the dependent variable OCB, the Kolmogorov-Smirnov statistics showed that the value (0.026) was significant at the 0.05 level, suggesting a violation of the assumption of normality (Appendix B, Table 10). However, there was a reasonably straight line in the normal probability plot, which signaled a normal distribution of the scores of OCB (Appendix B, Graph 14).
With respect to the mediating variable organizational commitment, the Kolmogorov-Smirnov statistics expressed that the values for affective commitment and
normative commitment were significant, violating the assumption of normality (Appendix B, Table 10). However, the normal probability plots for the two variables showed a reasonably
straight line, indicating normal distribution of the scores (Appendix B, Graph 19 and 20). The values for four types of the work values were significant, suggesting a violation of the assumption of normality. However, the normal probability plot showed a reasonably straight line in each plot of the scores of the moderating variables, signaling that the scores on intrinsic work values, extrinsic work values, social work values, and prestige work values
were normally distributed (Appendix B, Graph15, 16, 17 and 18).
Based on the ten boxplot graphs of the variables, (Appendix B, Table 11), only the mediating variable affective commitment did not have outliers. In the case of distributive justice, only one case was marked as an outlier – observation 134. Regarding procedural justice, two cases were identified as outliers – observation 74 and 115. With respect to interactional justice, three outliers were highlighted – observation 134, 172 and 225. The boxplot graph on OCB showed three outliers – observation 56, 76 and 224. Finally,
normative commitment had two outliers – observation 121 and 225 (Appendix B, Graph 21, 22, 23, 24, 29 and 30).
In the case of work values, intrinsic work values had five outliers – observation 114, 160,
232, 236 and 247. In the case of extrinsic work values, nine cases were considered as outliers
– observation 1, 2, 9, 37, 69, 141, 169 and 233. When social work values was the moderator, there was one outliers highlighted in the boxplot – observation 232. Finally, prestige work
values had six outliers – observation 20, 39, 223, 232, 236 and 247 (Appendix B, Graph 25, 26, 27 and 28).
However, those outliers were not excluded from the data set since the 5% trimmed mean values were relatively equal to the man values in the analysis (Appendix B, Table 9). The boxplot graph of extrinsic work values showed that there were two cases pointed out as extremes (indicated with an asterisk *), which are points extending more than three box-lengths from the edge of the box (Appendix B, Graph 26). However, there extreme cases were not removed from the data set since extrinsic work values do not affect any variables in this research.
5. Regression Analysis
In order to test the fourteen hypotheses developed in this study, multiple regression analyses were used since there were three types of relationships to be tested, direct, mediation and moderation. Two of the hypotheses test (H3a and H3b) for the mediating effect of affective commitment and normative commitment, three of the hypotheses test (H6a-c) for the moderating effects of three dimensions of work values, and the last nine hypotheses test for direct relationships. All regression analyses were conducted by using the data reported by the 250 participants.
Linear regressions were carried out to test for the direst relationships proposed in the nine hypotheses, the control variables age, gender, level of education and organizational
tenure were included, corresponding to Model 1 in the regression tables. In the second stage,
the independent variables were added, corresponding to Model 2.
The mediation effect of organizational commitment on the relationship between organizational justice and OCB was tested with model 4 of PROCESS while the variables age, gender, level of education and organizational tenure were included as control variables (Hayes, 2012b).
To test the moderation effect of different dimensions of work values on the relationship between organizational justice and OCB, model 1 of PROCESS was carried out for each of the three dimensions while age, gender, level of education and organizational tenure were included as control variables (Hayes, 2012b). Moderation analysis investigated whether different dimensions of work values can strengthen or weaken the relationship between organizational justice and OCB. In addition, interaction term examined whether the
hypotheses are supported or rejected.
With respect to Hypothesis 1, which hypothesizes that organizational justice has a
positive influence on OCB, the results are represented in Table 12 below. The results indicate
that 18.6% of the variance in OCB can be explained by the control variables. This model is significant at the 0.05 level (F = 13.950; Sig. =0.000). When the independent variable
organizational justice was added to the model, R square increased with 7% and 256.6% of
the variance of OCB can be explained. Model 2 is significant at 0.05 level (F Change
=23.063; Sig. F Change=0.000). The regression coefficient for organizational justice is
positive and significant (Beta =0.291; Sig. =0.000), resulting in the fact that H1 is supported. However, there are two variables in Model 1 that are significant at 0.05 level, including level of education (Beta =0.365; Sig. =0.000) and organizational tenure (Beta
=0.333; Sig. =0.000). The finding shows that education status and work experience were
positively related to engage in OCB.
Table 12 – Linear Regression between organizational justice and OCB
Variables Model 1 Model 2
Adj. R2 = .172 Adj. R2 = .241
Beta Sig. Beta Sig.
Age -.153 .070 -.060 .471 Gender -.074 .208 -.068 .228 Level of education .365 .000 .280 .000 Organizational tenure .333 .000 .248 .002 Organizational justice .291 .000 R square .186 .256 F 13.950 16.777 R square Change .186 .070 F Change 13.950 23.063 Sig. F Change .000 .000 N 250 250
The results are outlined in Table 13. The results show that 18.6% of the variance in OCB can be explained by the control variables. This model is significant at the 0.05 level (F = 13.950;
Sig. =0.000). When the variable affective commitment was added to the model, R square
increased with 0.1% and 18.7% of the variance of OCB can be explained. However, Model 2 is not significant (F Change =0.499; Sig. F Change=0.481). The regression coefficient for affective commitment is positive but not significant (Beta =0.045; Sig. =0.481), meaning that H2a is not supported.
Table 13 – Linear Regression between affective commitment and OCB
Variables Model 1 Model 2
Adj. R2 = .172 Adj. R2 = .171
Beta Sig. Beta Sig.
Age -.153 .070 -.151 .074 Gender -.074 .208 -.077 .190 Level of education .365 .000 .348 .000 Organizational tenure .333 .000 .325 .000 Affective commitment .045 .481 R square .186 .187 F 13.950 11.237 R square Change .186 .002 F Change 13.950 .499 Sig. F Change .000 .481 N 250 250
Hypothesis 2b to be tested was the relationship between normative commitment and OCB. The results are reported in Table 14. The results show that 18.6% of the variance in OCB can be explained by the control variables. This model is significant at the 0.05 level (F = 13.950;
Sig. =0.000). When the variable affective commitment was added to the model, R square
increased with 1.8% and 20.4% of the variance of OCB can be explained. Model 2 is significant (F Change =5.537; Sig. F Change =0.019). The regression coefficient for normative commitment is positive and significant at the 0.05 level (Beta =0.145; Sig.
