The Effects of Financial Distress and the use of Bonus Systems on Unethical Behavior of Employees
Master Thesis MSc. Business Administration – Strategy track Student: Jacco Pols 10870954 Jacco.pols@student.uva.nl Supervisor: J.W. Stoelhorst Date: June 24, 2016
Statement of originality
This document is written by Student Jacco Pols, who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents
Abstract
In the last couple of years many companies have received harsh criticism for the unethical decisions that were made by their employees. What drives those employees to make these unethical decisions is both theoretically and practically important for firms and their managers. This research shows that employees who perceive higher pressure to perform are more likely to engage in unethical behavior. Furthermore, drivers of this perceived pressure to perform of employees in a firm are the use of bonus systems and financial distress in the firm. In addition to this, the relationship between bonus systems and the perceived pressure to perform is moderated by the love of money of individuals. No support was found for the moderation of parochialism on the relationship between financial distress in the firm and perceived pressure to perform. For this study a total of 240 individuals participated in the survey in order to investigate the hypotheses. A vignette study was used to obtain the data needed for the t-tests and regression analysis.
Table of Contents 1. INTRODUCTION 1 2. LITERATURE REVIEW 3 2.1 DEFINITION OF ETHICAL DECISIONS 3 2.2 ANTECEDENTS OF ETHICAL DECISION-MAKING 5 2.3 PERSONALITY TRAITS AFFECTING ETHICAL DECISION-MAKING 6 2.4 SITUATIONAL FACTORS AFFECTING ETHICAL DECISION-MAKING 8 2.5 INTERACTION BETWEEN TRAITS AND SITUATIONAL FACTORS ON ETHICAL DECISIONS 10 2.6 RESEARCH ON ETHICAL DECISION MAKING IN FIRMS 12 3. THEORETICAL FRAMEWORK 14 3.1 THE CONCEPTUAL MODEL 14 3.2 HYPOTHESES RELATIONSHIP FIRM PERFORMANCE AND ETHICAL DECISION-MAKING 15 4. RESEARCH DESIGN 20 5. RESULTS 23 5.1 RELIABILITY ANALYSIS 24 5.2 DESCRIPTIVE STATISTICS 26 5.3 NORMALITY ANALYSIS 30 5.4 CORRELATIONS 33 5.5 INDEPENDENT SAMPLE T-TEST 37 5.6 REGRESSION ANALYSIS 39 5.6.1 ADDITIONAL REGRESSION ANALYSIS 47 6. DISCUSSION 49 6.1 RESULTS 50 6.2 IMPLICATIONS 52 6.2.1 THEORETICAL IMPLICATIONS 55 6.2.2 PRACTICAL IMPLICATIONS 56 6.3 LIMITATIONS AND AVENUES FOR FUTURE RESEARCH 58 6.4 CONCLUSION 60 REFERENCES 62 APPENDIX 73 APPENDIX A 73 APPENDIX B 79
1. Introduction
When and how employees make unethical decisions in times of prosperity and adversity is both theoretically and practically important for firms and their managers so they can prevent unethical behavior of their employees. Over the last couple of years lots of unethical decisions have come to light. For example Volkswagen, KPMG and others have received harsh criticism for decisions that the public judged to be unethical but that the employees made to improve either the companies’ and / or their own financial situation. What drove them to make these unethical decisions beforehand is unclear.
The ethical decision-making process is something that has been studied extensively over the last two decades (Craft, 2012; Ford & Richardson, 1994; O’Fallon & Butterfield, 2005). An ethical decision is a situation in which an employee must reflect upon competing moral standards and/or stakeholder claims in determining what is the appropriate decision or action (Kelman & Hamilton, 1989; Velasquez & Rostankowski, 1985). This would mean that ethical decisions have impact on others. For example, in a period where the company is performing poorly, a manager might have to decide if the personnel will receive a promised pay rise. Giving this pay rise as promised will be the right ethical decision for the personnel, however such a pay rise may endanger the firm further and may not be accepted by the shareholders of the company.
Researchers have investigated which dimensions of the organizational environment and which personality traits affect ethical decision-making (Hegarty & Sims, 1979; Marquardt, 2010; Ruedy & Schweitzer, 2011) and, more importantly, how the organizational environment can be changed by top
management so that ethical decision-making can be nurtured and unethical decisions will not harm the company and/or its employees in the future (Elango, Paul, Kundu, & Paudel, 2010; Ho, 2010; Stohs & Brannick, 1999). As a result, we already know that personal circumstances such as financial difficulties can affect ethical decision-making (Aronsson & Gustafsson, 2005; Honeycutt, Glassman, Zugelder, & Karande, 2001). Furthermore it has been found that setting group targets and bonuses increases the perceived pressure to perform of the group (Durham, Locke, Poon, & McLeod, 2000; Weldon & Weingart, 1993). This paper examines whether these relationships extend to the case of employees making ethical decisions on their firm’s behalf through a higher perceived pressure to perform on the employees.
In other words, this paper looks at the relationships between firm performance and the use of bonus systems and unethical behavior of employees through perceived pressure to perform, and at what is moderating this relationship. Performance in this research is defined as the result of revenue minus expenses. If the result is positive there is good performance and when the result is negative there is bad performance. With bad performance there will be financial distress and this may affect the employees when making decisions. A bonus system is used when employees are rewarded when they fulfill targets.
This research will contribute to the literature in the following way. The paper will build on findings of ethical decision-making at an individual level and investigate if these findings also hold in a work environment and how they may differ. In particular, while it is established that self-interest and personal financial distress affect individuals’ ethical decision-making outside of the workplace (Brenner & Molander, 1977; Tang & Chiu, 2003) it is still unclear how
a firm’s financial distress would play out as a contextual influence on ethical behavior of employees when making decisions on behalf of the firm. Therefore this research will look at whether there is a relationship between performance of the company and the ethical decision making of managers of these companies. Further, the study will investigate whether applying a pay for performance systems also affects the ethical behavior of employees through a higher perceived pressure to perform.
The structure of this paper will be as follows. Firstly, definitions of ethical decisions will be given. Secondly, Rest’s model of ethical decision-making will be discussed. This model talks about how ethical decisions are made. Thirdly, factors affecting ethical decisions that have been found in the literature are discussed. Fourthly, the research gap and research questions are discussed together with a model of the research. In the final sections the result of the analysis and the discussion section with implications and limitations are found.
2. Literature review
2.1 Definition of ethical decisions
According to previous literature it is hard to state an uniform definition of an ethical decision (Beauchamp & Bowie, 1979; Cavanagh, Moberg, & Velasquez, 1981; Jones, 1980). Several authors, such as Trevino (1986) and Hunt & Vitell (1986) even write about ethical dilemmas but do not provide definitions for ethical and unethical behavior at all.
This paper defines an ethical issue as a problem where a person must reflect upon competing moral standards and/or stakeholder claims in determining what is the appropriate decision or action. This follows from the
definitions of ethical decisions and ethical issues. Kelman & Hamilton (1989) define an ethical decision as a decision that is morally accepted by others, and on the other hand, an unethical decision is either illegal or not morally accepted by others. An ethical issue is an issue where the person’s decisions may affect others positively or negatively (Velasquez & Rostankowski, 1985). In the definition given by this article the decision maker thus knows he or she is facing an ethical issue (and thus is affecting others), and that the community could judge him or her on his actions.
According to Rest (1986) ethical decision making can be seen as a process of four steps that need to be fulfilled before the decision is made. These four steps are: 1) moral awareness; 2) moral judgment; 3) moral intent; and 4) moral behavior. In brief, moral awareness is recognizing that the issue is an ethical issue or has some ethical sides. Moral judgment is the ability of the decision maker to decide whether the action is ethically responsible (Craft, 2012). Moral intent is the ability for the decision maker to place the moral concern above other concerns when looking at the ethical issue. Moral behavior is the ethical behavior itself, where either something or nothing can be done.
Rest (1986) states each step in the ethical decision process is different and that the completion of one step does not always lead to the completion of the following step. For example, a manager who knows whether his decision is ethical or not does not automatically place the moral concern above other concerns when looking at the ethical issue. This could lead to making unethical decisions. Fassin (2005) found that managers have different reasons to place other concerns above the moral concern when making a decision. Reasons for
this can be pressure from shareholders, evolution of society, globalization of the economy, short-term tactics or dominance of financial consideration.
2.2 Antecedents of ethical decision-making
From a person-situation interactionist perspective (e.g. Trevino 1986), antecedents of ethical decision-making can be classified in two categories that interact with each other: individual differences (personality traits) and situational factors. With regard to individual differences, the recent empirical work by Cohen, Turan, Morse, & Kim (2014) supports that it is important to take into consideration individual differences to explain unethical behavior. They present a tripartite model for understanding moral character (i.e. “an individual’s disposition to think, feel, and behave in an ethical versus unethical manner” (Cohen et al., 2014)) in which they include motivational, ability and identity elements. Consideration of others is the motivational element, self-regulation is the ability element and the identity element is moral identity. Their research indicates that the impact of moral character is significant on work outcomes and thus affects the ethical decision-making processes.
Another important facet of the ethical decision-making process is that individuals must perceive the situation as an ethical problem (moral awareness) (Butterfield, Trevin, & Weaver, 2000). Singhapakdi, Rao, & Vitell (1996) found individuals who perceive the situation as having an ethical element are more likely to form an ethical judgment. However Valentine & Fleischman (2003) found recognizing the ethical element in the decision was not related to forming an ethical judgment. This would suggest that even if the individual does recognize there is an ethical element, Rest’s (1986) ethical decision-making
process could be interrupted, as there is no ethical judgment made by the decision maker.
Empirical literature has, by far, focused mostly on individual factors affecting ethical decision-making by individuals (Ford & Richardson, 1994). This paper will look whether there is a positive relationship between financial distress of the firm and having bonus systems and unethical decision of employees through a higher perceived pressure to perform and what factors are moderating this relationship. Hence, especially the findings of Brenner & Molander (1977) and Weldon & Weingart (1993) are interesting, they found having financial distress at home increases the perceived pressure to perform at the organizational context and setting group targets increases the perceived pressure to perform of the group. As stated, it is important to separate individual factors that are influencing ethical decision making into traits and situations. Traits refer to the personality of the individual that is making an ethical decision. Situation refers to the situation in which the ethical decision has to be made.
2.3 Personality traits affecting ethical decision-making
One of the individual factors that existing literature has found to have a significant effect on ethical behavior is gender. In general it has been found that women behave more ethically than men do. For example, Eweje & Brunton (2010) found that females are more ethically aware than males, Zopiatis & Krambia-Kapardis (2007) found females are more ethical than males in a study of perception and Glover, Bumpus, Logan, & Ciesla (1997) found women make more ethical decisions than men. It is interesting to see if this also holds in the case of making ethical decisions on firm’s behalf. It might be interesting to look
whether firm performance influences this finding and whether females still are more ethical when there is financial distress or when a bonus system is used.
Another individual factor the existing literature has found to have a significant effect on ethical behavior is age, however there are some mixed findings on this factor. For example, Elm & Nichols (1993) and Sankaran & Bui (2003) found that age was negatively related to moral reasoning and the older the individual becomes, the less ethical he/she becomes. This would mean that younger individuals would be more ethical than older individuals, however Hunt & Jennings (1997) and Kim & Chun (2003) found that younger generations gave less ethical responses and younger teams tended to make the most unethical decisions. In general, it can be stated the results have been mixed and therefore it is hard to state whether young or old individuals are more ethical. Therefore, it is difficult to state whether or at least how age moderates the relation between performance and ethical decision-making in general. A final interesting factor that affects ethical decision-making is the years of work experience the decision maker has. Eweje & Brunton (2010) found more experienced employees appeared to be more ethical, McCullough & Faught (2005) have shown work experience was positively related to a persons tendency to be more moralistic and Valentine & Rittenburg (2006) show increased age and work experience have a positive relationship with ethical judgment and ethical intentions of individuals. Furthermore, according to Ruegger & King (1992) older and thus more experienced students are more ethical than less experienced students. An explanation given by Knotts, Lopez, & Mesak (2000) is the old expression that wisdom comes with age, and thus individuals behave more ethical when they get older. Another explanation for the
positive relationship between years of work experience and ethical behavior is given by Cron (1984). He introduced a framework to organize ideas about workforce development with four stages: exploration, establishment, maintenance and disengagement (Weeks, Moore, McKinney, & Longenecker, 1999). The exploration stage is about acceptance from peers and superiors. If there are high demanding standards, the relative new employee might engage in unethical behavior due to the pressure to perform. The establishment stage is according to Cron (1984) a competitive stage and employees are concerned with superior performance. While the pressure to perform can be more internal than from a manager, the high need for performance may lead to ethical compromise (Weeks et al., 1999, p. 305). In contrast, the maintenance and disengagement stages are characterized by lower pressure and more focus on helping others. In these stages there is less temptation to behave unethical because of higher degrees of job security. Weeks et al. (1999) therefore state when an employee experiences a high need for performance this is likely to increase his engagement in unethical behavior.
2.4 Situational factors affecting ethical decision-making
Situational factors can also affect the ethical decision-making process. Situational factors can be divided in issue characteristics and contextual factors. Jones (1991) talks about issue characteristics that affect the ethical decision-making process. For each decision the consequences (i.e. magnitude, probability, temporal immediacy and concentration of the effect) of the action affect the decision that is made. Further the social consensus that a proposed act is right or wrong and the proximity or the feeling of nearness all affect the decision that will be made in the end. Contextual factors can be organizational infrastructure, such
as formal ethic codes and policies, communication, training, monitoring systems, sanctions, rewards, ethical climate and organizational culture (Oldham & Cummings, 1996).
Honeycutt et al. (2001) studied how income affects the ethical behavior of individuals. They found people with a fixed income are behaving more ethical than people without a fixed income or with a pay-for-performance compensation method. This would mean that people without a fixed income, or people whose income relies on pay-for-performance would behave less ethical than those with fixed income jobs. People with less income will experience more financial distress and might be more eager to behave unethically.
Looking at how salespeople make ethical decisions, Verbeke, Ouwerkerk, & Peelen (1996) found that contextual factors play an important role when making an ethical decision. Internal communication and the choice of a control system especially affected ethical decision-making in their study. Other factors can be group influence and personal situation, for example stress. Verbeke et al. (1996) also found that the management of a firm can influence ethical decision-making of individuals in the firm. This can for example be done by using goal setting theory, as long as it is implemented in the right way (Ordóñez, Schweitzer, Galinsky, & Bazerman, 2009) or by investing in human resource management (Delaney & Huselid, 1996).
An important finding regarding situational factors is the one of Leitsch (2004), who found that the type of unethical situation (approving a questionable expense report, manipulating company book, violating company policy and extending questionable credit) influenced accounting students’ perception of the moral intensity and their moral judgment. This means that in different
circumstances students perceive the ethical issue differently than under other circumstances. In addition, Church, Gaa, Khalid Nainar, & Shehata (2005) found that personal gain due to misrepresentation was significantly positive related to unethical behavior. Further, Schweitzer & Gibson (2007) looked at explanations for unethical behavior for individuals. They found individuals were significantly more likely to engage in unethical behavior when they perceived that they had been the targets of “unfair” behavior. Participants in this unfair condition viewed the use of unethical behavior as more justifiable and they stated they would behave unethically more likely again. Specifically, when individuals perceived they had been treated unethically, they felt happier and less guilty, when they behaved unethical themselves (Schweitzer & Gibson, 2007). This would mean that if individuals perceive their (negative) individual situation is caused by unfair behavior of others they will behave less ethical themselves. This is also supported by Street & Street (2006) who found individuals with an external locus of control were significantly more likely to select the unethical option than those with an internal locus of control.
2.5 Interaction between traits and situational factors on ethical decisions
Morris, Marks, Allen, & Peery (1996) found personal values, such as sense of accomplishment and responsibility, are negatively related with unethical behavioral intentions, meaning that if people know that a certain behavior is unethical they will not engage in it. However, Vinokur & Van Ryn (1993) found that when people are unemployed and thus might experience financial difficulties, they will loose some of their values and might start behaving unethical in order to survive. This would suggest that situational factors override
personality traits regarding ethical decision making of individuals. An individual that normally is making ethical decisions thus might behave unethical in certain situations. On the other hand, it also occurs that individuals sacrifice their own self-interest for the group to which they belong. They even do so when there is no net-benefit for themselves or the group (Baron, 2001). The tendency of people to favor a group at the expense of their own self-interest is called parochialism (Schwartz-Shea & Simmons, 1991). A closely related research tradition concerns “in-group bias” (Brewer, 1979). Sherif, Harvey, White, Hood, & Sherif (1961) give an explanation for parochialism and in-group bias. They state people’s own interest are mobilized when their group is in competition and their competitive behavior actually is rationally self-interested as this might be needed for survival of the group. This would mean that individuals are likely to choose the groups interest above their own when the group is in fierce competition. Another explanation for parochialism is given by public choice theory and rational choice theory. Hidden in these explanations is the assumption that people go beyond their self-interest to act on behalf of the group due to self-interest illusion applies more to groups with which people identify (Brennan & Buchanan, 1985; Green & Shapiro, 1995). This would mean that people “forget” their own self-interest while making decisions for a group. An interesting addition to this is given by Vinokur & Van Ryn (1993), who state that individual values may be dropped due to situational factors. An individual that normally sacrifices his self-interest for the group might change this behavior due to circumstances, such as financial difficulties.
2.6 Research on ethical decision making in firms
Empirical literature has focused less on ethical decision-making in firms. Those researches that focus on ethical decision-making in firms mostly look at the consequences of unethical behavior by firms on consumer behavior, which are all negative. The following papers look at reasons why employees would behave unethically when acting on behalf of the firm.
Vardi & Wiener (1996) look at factors that affect misbehavior in organizations. They state individuals differ in their tendency to misbehave while in organizations. Factors that affect this are personality, person-organization value congruence, generalized value of loyalty and duty, personal circumstances and dissatisfaction of personal needs by the organization. Furthermore Vardi & Wiener (1996) state that there also is a difference in contextual conditions in organizations. Factors that affect either more or less misbehavior of personnel in the organization are opportunities to misbehave, control systems, culture in the organization, cohesiveness in the organization and organizational goals.
Another interesting contribution has been made by Nkundabanyanga, Mpamizo, Omagor, & Ntayi (2011). They found that if employees are willing to sell unprofessional assignments that do harm to others for monetary benefits, or if they are willing to succeed regardless on the way they reach these successes, this will lead to increased unethical behavior. This corresponds to Tang & Chiu (2003), which state that the love of money is the cause of unethical behavior. Love of money is an individual’s personalized attitude towards money (Tang, Tang, & Luna-Arocas, 2005). Another finding by Nkundabanyanga et al. (2011) is that there is a positive relationship between impossible targets combined with fixed deadlines and engagement in unethical behavior of the employees. In
general, it is stated that compensation methods that rely on commissions and bonuses are likely to increase unethical behavior. A reason for this is these methods put pressure on the employees to get results (Román & Munuera, 2005; Tang, Chen, & Sutarso, 2008).
All taken together it can be concluded there are different reasons why individuals would behave unethically. Factors that affect this relationship can be personality traits, situational factors or an interaction between both. An important reason to behave unethical can be that this behavior is needed to survive or achieve strict targets. In general it can be stated there is a positive relationship between bad performance and unethical behavior, meaning that when performance is low people will behave more unethical. A influential factor in this relationship in the private situation can be perceived fairness (Church et al., 2005; Schweitzer & Gibson, 2007). Further, opportunities to misbehave, the work culture, organizational goals and control systems affect misbehavior of personnel in an organization (Vardi & Wiener, 1996). Other moderators can be the love of money (Tang & Chiu, 2003), parochialism (Schwartz-Shea & Simmons, 1991) and impossible targets combined with strict deadlines (Nkundabanyanga et al., 2011; Román & Munuera, 2005; Tang et al., 2008) and whether a bonus system is used (Román & Munuera, 2005; Tang et al., 2008). Furthermore, it is interesting that the engagement in unethical behavior differs between male and female (Eweje & Brunton, 2010; Glover et al., 1997; Zopiatis & Krambia-Kapardis, 2007).
3. Theoretical framework
3.1 The conceptual model
As the literature shows, quite some results have been found on the relationships between unethical behavior and personality traits, situational factors and an interaction between those. In particular, while it is established that self-interest and personal financial distress affect individuals’ ethical decision-making outside of the workplace, it is still unclear how firm performance and bonus systems would play out as a contextual influence on ethical behavior of employees when making decisions on behalf of the firm. Therefore this research will look at whether there is a relationship between financial distress of the company and whether or not a bonus system is used and the unethical decision making of the employees of these companies. Further the study will look at two personality traits as possible moderators of this relationship; love of money and parochialism.
The research question in this thesis therefore will consist of the following four questions:
1) Does financial distress of the firm affect the ethical decisions made by employees on their firm’s behalf through higher perceived pressure to perform? 2) Is this relationship moderated by parochialism? 3) Does using bonus systems affect the ethical decisions made by employees on their firm’s behalf through higher perceived pressure to perform? 4) Is this relationship moderated by love of money?
The moderators that will be looked into are parochialism and love of money. The mediator in this research will be the perceived pressure to perform of the employees, which is hypothesized to be higher when firms have financial distress or are using a bonus system. The model of the research can is as follows: Figure 1: conceptual model firm performance and bonus system on unethical behavior 3.2 Hypotheses relationship firm performance and ethical decision-making In the literature review above, various drivers of ethical and unethical decision making have been identified. In this section these findings will be linked to the organizational context. It is unclear if financial distress and using bonus systems affect the ethical decision-making of individuals in a firm through higher perceived pressure on the employees.
Altman (1983) shows managers have higher demands of employees when the firm experiences financial difficulties. Companies with financial distress might face bankruptcy or reorganizations and this is likely to increase Financial distress of the firm Unethical behavior of employees Bonus system Perceived pressure to perform Love of money Parochialism + + + +
the perceived pressure to perform of the employees, as they don’t want this. Furthermore, Brenner & Molander (1977) found an individual’s personal financial situation is one of the six reason why people experience pressure and start behaving unethical and Aronsson & Gustafsson (2005) found the financial situation decreases absenteeism on the work floor due to increased pressure to perform. Whether the relationship found between financial distress in the personal situation and perceived pressure to perform extends to a relationship between financial distress in the organizational context and perceived pressure to perform is unclear. However, when a firm is experiencing financial difficulties it is likely that the employees perceive a higher pressure to perform as well. The first hypothesis thus will be:
H1. Financial distress in the firm has a positive relationship with perceived pressure to perform of the employees
When competition is fierce, managers may use a pay for performance system in order to increase performance, and when employees reach those goals they receive a monetary bonus. Linderman, Schroeder, Zaheer, & Choo (2003) state in general hard goals increase performance, but when the goals become impossible this leads to a drop in performance. Furthermore, in general it is stated that compensation methods that rely on commissions and bonuses are likely to increase pressure on the employees to get results (Román & Munuera, 2005). Weldon & Weingart (1993) state setting group goals increases the pressure to perform on the individuals in the group. Furthermore, Durham, Locke, Poon, & McLeod (2000) found group goals could decrease performance and increase
performance pressure. These finding all suggest using a bonus system in the company is likely to make employees of the firm experience a higher perceived pressure to perform. Therefore the second hypothesis will be: H2: Bonus systems have a positive relationship with perceived pressure to perform of the employees
Firms that are in financial distress need to increase their performance. A common method to increase performance is by setting goals for the employees (Locke, Shaw, Saari, & Latham, 1981). The combination of goal setting with using a pay for performance systems increase the pressure to perform on employees on the group level (Durham et al., 2000; Román & Munuera, 2005; Weldon & Weingart, 1993). Altman (1983) already showed that managers have higher demands of their employees when the firm has financial difficulties. Furthermore, Brenner & Molander (1977) and Aronsson & Gustafsson (2005) found a persons’ personal financial situation influences the perceived pressure to perform at his work. Altogether, these findings would suggest that when managers use a pay for performance systems and when the firm is experiencing financial difficulties this will increase the perceived pressure to perform of the employees even further. A combination of both financial distress and having bonus systems in the firm thus are likely to increase perceived pressure to perform of the employees. The third hypothesis therefore will be:
H3: There is a positive interaction of the effects of financial distress and the use of bonus systems on perceived pressure to perform
Bekkers & Wiepking (2010) state individual don’t have a normal tendency to help others. However, Baron (2001) mentions individuals may sacrifice their own self-interest for the group to which they belong too when the groups need their help. An explanation for this is people go beyond their self-interest to act on behalf of the group because their self-interest makes place for group-interest (Brennan & Buchanan, 1985; Green & Shapiro, 1995). The tendency of individuals to favor the group to which they belong at the expense of their own self-interest is called parochialism (Schwartz-Shea & Simmons, 1991). Brenner & Molander (1977) showed financial distress lead to higher perceived pressure and financial distress can be seen as a legitimate reason for helping the group (Bickman & Kamzan, 1973; Taormina & Messick, 1983). It therefore is interesting to investigate whether individuals sacrifice their own self-interest when the company they belong to is experiencing financial difficulties. Altogether, these findings lead to a need to investigate whether parochialism influences the relationship between financial distress of the firm and perceived pressure to perform. The fourth hypothesis therefore will be:
H4: The positive relationship between financial distress of the firm and perceived pressure to perform is moderated by parochialism, so that this relationship is stronger for higher parochialism
According to Tang & Chiu (2003) the love of money is the root of all unethical behavior in the business context. Omagor, Mubiru, & Nkuutu (2009) state this love of money is associated with success. As money is viewed as an indicator of
success this will make people obsessed with money to prove their success (Furnham & Argyle, 1998). Money thus can motivate people to reach goals when money is rewarded. This implies that people will work harder when they are paid on commission basis with goal setting. The fact that money is needed to show success is likely to put pressure on employees to earn more money, especially when they are paid for their performance. It is likely that this pressure is higher for those with high love of money (Furnham & Argyle, 1998). As Durham et al. (2000) and Weldon & Weingart (1993) already found group goals increases the perceived pressure to perform of the group it is likely that the love of money strengthens this relationship and increases the perceived pressure even more. These findings lead to a need to investigate whether love of money influences the relationship between bonus systems and unethical behavior of employees. The fifth hypothesis therefore will be:
H5. The positive relationship between bonus systems in the firm and perceived pressure to perform is moderated by love of money, so that this relationship is stronger for higher love of money
Nkundabanyanga et al. (2011) found that setting impossible targets with fixed deadlines will increase unethical behavior in firms. A reason for this is given by Weeks et al. (1999) who states a high need for performance can lead to ethical compromises. As mentioned above, using goal setting for the group increases the pressure to perform of the employees (Durham et al., 2000; Weldon & Weingart, 1993). Román & Munuera (2005) and Tang et al. (2008) state increased pressure to perform is likely to increase unethical behavior because employees will do
anything to get results and there are several drivers of perceived pressure to perform. Furthermore Vinokur & Van Ryn (1993) showed personality traits can be overruled with certain situational factors, such as increased pressure to perform. This would suggest when employees perceive more pressure to perform they are more likely to drop some of their ethical values and start behaving unethically. Altogether, these findings suggest that when employees perceive they have a higher pressure to perform they are more likely to behave unethical. The sixth and final hypothesis thus will be:
H6. Perceived pressure to perform has a positive relationship with unethical behavior of employees
4. Research design
The research used in this study consists of two different instruments. First, an experimental vignette is used to manipulate the financial situation and the reward system of the firm. Included in this vignette are scales measuring how much pressure to perform employees perceive and how unethical they would behave in the outlined situation. Furthermore, manipulation check questions were included to test whether the respondents understood the situation and if the vignettes are realistic. Secondly, a survey was included with the vignette in order to measure two personality traits, parochialism and love of money, and personal information about the respondents’ age, level of education and gender. For this research, a between-subject design is used. This allows the research to split up the respondents into four groups that are presented one of the four (two by two) different vignettes. This is done to measure the influence
of the independent variables on the mediator and dependent variable and be able to compare this across the different groups. The different versions of the vignettes are based on a study found in Smith, Skalnik, & Skalnik (1999) and can be found, together with the full set of questions, in appendix A.
The vignettes are used to test the hypotheses as stated in the theoretical framework. Vignettes are built on components of classical experiments as well as surveys (Aguinis & Bradley, 2014; Atzmüller & Steiner, 2010). By doing so vignettes eliminate some weaknesses of the two approaches. Surveys have a high external but low internal validity and experiments a low external but high internal validity (Atzmüller & Steiner, 2010). It is important that the same context as in real life is created in the vignette, and if this is not possible other techniques should be used (Aguinis & Bradley, 2014; Lohrke, Holloway, & Woolley, 2010). After the vignettes were developed they were pre-tested on a selected group of colleagues, friends and relatives. The pre-test results and comments that were obtained were used to improve the vignettes and create a real life experience. The vignettes were written very concise in order to make it easy for the respondent to answer all questions in a short timeframe, while still being concentrated to do so correctly.
For each vignette the respondent was asked to what extend he agrees with statements regarding the (un)ethical behavior that he or she will show and the pressure to perform he or she experiences. This was assessed on a 5-point Likert scale with the endpoint strongly disagree (1) and strongly agree (5). Since no scales exist to measure the perceived pressure to perform or the behavior someone shows they were developed for the use of this study. The statements to measure the unethical behavior are as following: 1) I would inform the customer
about the specifications (reversed), 2) I would not say anything about the specifications to make sure that the customer buys the product, 3) Making the sale is more important for me than informing the customer about the specifications. The statements to measure the perceived pressure to perform are as following: 1) I experience a high pressure to perform, 2) A lot depends on my performance at company X, 3) The pressure to perform at company X is high for me. The full set of question can be found in appendix A.
In order to measure the moderators in this study, parochialism and love of money, existing scales were adopted. Respondents were asked to fill in the horizontal and vertical collectivism measurements in order to assess parochialism (Hofstede, 2001; Triandis & Gelfand, 1998). This scale uses a 9-point Likert scale with endpoints strongly disagree (1) and strongly agree (9). In order to measure the love of money the love of money scale was adopted from Tang et al. (2006). This measurement uses a 5-point Likert scale with endpoints strongly disagree (1) and strongly agree (5). To measure the extent to which the respondents understood the vignette and if the manipulations worked two control questions were asked. In order to control whether the respondents understood the financial situation of the company they were asked whether the financial situation of company X is good and answer this question with true or not true. To control the manipulation with the bonus system respondents were asked to indicate whether they could earn more when they sell more and answer this question with true or not true. Finally respondents were asked to rate the extent to which they found the vignette realistic. This was assessed on an 11-point Likert scale with endpoints not realistic at all (1) and fully realistic (11).
A total of 274 participants started the survey, and of these 240 respondents fully completed it while also answering the control questions correctly, which corresponds to a dropout rate of 12,4% in total. Of these 34 filtered responses 28 (10.2%) did not finish the survey completely and 6 (2.2%) did not answer the control questions correctly. 5. Results The following section reports the results of this research. First, it was necessary to compare the data obtained from the responses in Dutch and the responses in English. In order to do so a t-test was done for the perceived pressure to perform and the unethical behavior of employees. The results can be found in tables 1 and 2 below. The differences between the means are .0772 for perceived pressure and .0753 for unethical behavior, which is quite small. As Levene’s test gives p > 0.05 equal variances are assumed for both perceived pressure to perform and unethical behavior. The p-value of .079 for pressure to perform and .092 for unethical behavior of employees indicate there are no differences between the groups. Therefore all data was merged and used for the analysis. Table 1 – Independent sample t-test Dutch and English respondents Language N Mean Std. Deviation Perceived pressure to perform Dutch 172 3,6252 .82275 English 68 3.7024 .69412 Unethical behavior of employees Dutch 172 2.2451 .82417 English 68 2.3204 .91311
Table 2 – Levene’s test for equal variances perceived pressure and unethical behavior. Levene’s test for equal variances T-test F Sig. Sig (2-tailed) Mean difference Perceived pressure to perform Equal variances assumed 2.276 .133 .079 -.0772 Equal variances not assumed .071 -.0772 Unethical behavior of employees Equal variances assumed 1.352 .246 .092 -.0753 Equal variances not assumed .074 -.0753
Secondly, the results of the reliability analysis for the variables that are measured with multiple items are reported. The Cronbach’s alpha coefficient was analyzed for unethical behavior of employees, love of money, parochialism, and for perceived pressure to perform. Thirdly, descriptive statistics are given for the categorical and the continuous variables of the study to provide an overview of the data. As a next step the results of the normality analysis are outlined, in order to ensure the normal distribution of the scores on the dependent variable in each of the four vignette cases. Next, a correlation analysis is carried out and the most significant correlations are reported. Finally, t-tests and hierarchical multiple regressions were performed in order to test the hypotheses that were formulated in the theoretical framework of this paper.
5.1 Reliability analysis
For the purpose of this research, four variables were measured using multiple questions: the dependent variable, the mediator variable and the moderating variables.
The dependent variable unethical behavior of employees was captured using 3 questions and tested under the manipulation of the use of bonus systems and financial distress in the firm. The Cronbach’s alpha of unethical behavior was .747 (Table 3 – Reliability Statistics), which was above 0.7. This indicates that the variable had good internal consistency (Field, 2013). Furthermore, no negative values were registered in the Inter-item Correlation Matrix, meaning that all 3 items were always measuring the same underlying characteristic.
In the case of the mediating variable – perceived pressure to perform, the Cronbach’s alpha coefficient of the 3-item scale used to capture each of the variables, indicated an internal consistency of .779 (Table 3 – Reliability Statistics), indicating a good internal consistency (Field, 2013). Furthermore all items were measuring the same underlying characteristic as no negative values are found in the Inter-item Correlation matrix.
In the case of the moderating variables – love of money and parochialism, the Cronbach’s alpha coefficients of the scales used to capture the variables indicated an internal consistency ranging from .718 to .781 for the different love of money factors and .877 for all 9 items total. For parochialism the factors indicated an internal consistency ranging from .724 to .769 for the collectivism scales, and .788 for all 8 factors total. This thus indicates a good internal consistency (Field, 2013).
In conclusion, all multi-item variables had good internal validity, which signifies that they are uni-dimensional and that they can be recoded into individual variables.
Table 3 - reliability statistics
Variable Level Cronbach’s
Alpha Cronbach’s alpha based on standardized items N Unethical behavior of employees Total .747 .755 3 Perceived pressure to perform Total .779 .790 3 Love of money Total Rich Motivator Important .877 .741 .781 .718 .877 .751 .779 .772 9 3 3 3 Parochialism Total Horizontal collectivism Vertical collectivism .788 .724 .769 .796 .742 .769 8 4 4 5.2 Descriptive statistics In this study three categorical variables were used, namely the different vignette scenarios, gender and level of education. A description of all categorical variables can be found in table 4 – frequencies for categorical variables.
From a total of 240 respondents, 63 filled in vignette 1, containing financial distress and a bonus system, 57 filled in vignette 2, containing no financial distress and no bonus system. 59 respondents filled in vignette 3, containing no financial distress and a bonus system, and 61 respondents completed vignette 4, containing financial distress and no bonus system.
In terms of gender and level of education the survey is completed by 155 females (64.6%) and 85 males (35.4%). 32 (13.3%) respondents completed secondary school as highest education, 11 (4.6%) completed vocational training, 141 completed a bachelor study (58.8%), 55 (22.9%) completed a master study and 1 (0.4%) respondent clicked other as highest completed education.
At the beginning of the survey respondents were asked whether they would take the survey in Dutch or English. 171 (71.3%)respondents chose to fulfill the survey in Dutch while 69 (28.7%) respondents chose English.
Table 4 - frequencies for categorical variables
Variable Level Frequency % % valid Cumulative %
Vignette scenario 1 – Bonus, distress 2 – no bonus, no distress 3 – bonus, no distress 4 – no bonus, distress 63 57 59 61 26.3 23.8 24.6 25.7 26.3 23.8 24.5 25.8 26.3 50.1 74.6 100 Gender 1 – Male 2 – Female 85 155 35.4 64.6 35.4 64.6 35.4 100 Education 1 – Secondary School 2 – Vocational Training 3 – Bachelor 4 – Master 5 – Other 32 11 141 55 1 13.3 4.6 58.8 22.9 0.4 13.3 4.6 58.8 22.9 0.4 13.3 17.9 76.7 99.6 100 Language 1 – Dutch 2 – English 171 69 71.3 28.7 71.3 28.7 71.3 100
Next to the categorical variables, five continuous variables were used. The dependent variable, unethical behavior of employees in each of the scenarios, the mediator, perceived pressure to perform in each of the scenarios, the moderators, parochialism and love of money, and the control variable age. A description of all continuous variables can be found in table 5 – descriptive statistics for continuous variables. The continuous variables were measured using both 5 and 9-point Likert scales and respondents were asked to fill in their age. For all variables summary statistics, such as mean, median and standard deviation are given (Pallant, 2007).
The results for unethical behavior of employees, varied across the different scenarios. In total for all respondents the mean was 2.2903, the standard
deviation .85678 and values ranged form 1 to 5. In the scenario with bonus systems and financial distress the mean was 2.4021, the standard deviation 1.00207 and values were ranging from 1 to 5. In the scenario with no bonus and no financial distress the mean was 2.0468, standard deviation .69133 and values ranged from 1 to 4. In the scenario with bonus systems and no financial distress the mean was 2.2542, standard deviation .71207 and values ranged from 1 to 4.33. In the final scenario, no bonus systems and financial distress the values ranged from 1 to 5, the mean was 2.4372 and standard deviation .92604. Interestingly, the mean was highest in the situation with financial distress and no bonus system. Further the mean was higher in the situation with both financial distress and bonus system. In the situation with bonus systems and no financial distress the mean was lower than the average, which is quite interesting as there is hypothesized this will lead to higher perceived pressure to perform.
The results for perceived pressure to perform varied across the different scenarios. The average of all respondents was 3.6639, the standard deviation .80675 with values ranging from 1 to 5. In the scenario with bonus systems and financial distress the mean was the highest with 4.0159, the standard deviation .65700 and values were ranging from 2 to 5. In the scenario with no bonus and no financial distress the mean was 3.0292, standard deviation .86895 and values ranged from 1 to 4. In the scenario with bonus systems and no financial distress the mean was 3.8023, standard deviation .59086 and values ranged from 2.33 to 5. In the final scenario, no bonus systems and financial distress the values ranged from 1.67 to 5, the mean was 3.7596 and standard deviation .75063. The mean was higher for all situations where there was either financial distress or bonus
systems and only lower when people were paid a fixed salary and there was no financial distress.
In terms of parochialism and love of money values are varied. For horizontal collectivism values varied between 4.5 and 9, with a mean of 7.3379 and a standard deviation of .89561. For vertical collectivism values varied between 4 and 9, with a mean of 7.1391 and a standard deviation of 1.12960. The total mean of parochialism was 7.2385 with a standard deviation of .85442. The love of money scale differentiates between rich, motivator and importance of money. All levels have values ranging from 1 to 5. Rich has a mean of 3.7308, a standard deviation of .66777, motivator has a mean of 3.7294 and a standard deviation of .67727 and importance of money has a mean of 3.7503 and a standard deviation of .60920. The total mean for love of money was 3.7369 and the standard deviation .56764.
The average age of the respondents is 24.371 years old. The standard deviation of age is 4.76481 years and the values are ranging form 17 years old to 57 years old.
Table 5 – descriptive statistics for continuous variables
Variable Level N Min Max Mean Standard
Deviation Unethical behavior of employees 0 – Total 1 – Bonus, distress 2 – no bonus, no distress 3 – bonus, no distress 4 – no bonus, distress 240 63 57 59 61 1 1 1 1 1 5 5 4 4.33 5 2.2903 2.4021 2.0468 2.2542 2.4372 .85678 1.00207 .69133 .71207 .92604 Perceived pressure to perform 0 – Total 1 – Bonus, distress 2 – no bonus, no distress 3 – bonus, no distress 4 – no bonus, distress 240 63 57 59 61 1 2 1 2.33 1.67 5 5 4 5 5 3.6639 4.0159 3.0292 3.8023 3.7596 .80675 .65700 .86895 .59086 .75063
Parochialism 0 – Total 1 – horizontal collectivism 2 – vertical collectivism 240 240 240 4 4.5 4 9 9 9 7.2385 7.3379 7.1391 .85442 .89561 1.12960 Love of money 0 – Total 1 – rich 2 – motivator 3 - important 240 240 240 240 1 1 1 1 5 5 5 5 3.7369 3.7308 3.7294 3.7503 .56764 .66777 .67727 .60920 Age 240 17 57 24.371 4.76481 5.3 Normality analysis
A normality test was conducted to assess whether the variables perceived pressure to perform, unethical behavior of employees, parochialism and love of money were normally distributed.
With respect to perceived pressure to perform, the mediating variable had a mean value of 3.6639 and a 5% trimmed mean of 3.7022 (Appendix B, Table 21 – Normality Analysis – descriptives). This indicated that the more extreme scores do not have a substantial influence on the mean. Further, the negative skewness and positive kurtosis indicates the score of the variable are clustered towards the right end side of the graph and that the distribution is rather peaked in the center with long tails and eventually some outliers. (Appendix B, table 21 and figure 2)
In the case of the dependent variable unethical behavior of employees mean value was 2.2903 and a 5% trimmed mean of 2.2315 indicating that the most extreme score do not have a substantial impact on the mean. Furthermore, the positive skewness and kurtosis indicates that the distribution of the variable is clustered towards the left end side of the graph and that the distribution is rather peaked in the center with long tails (Appendix B, table 21 and figure 3).
With respect to love of money, the moderating variable had a mean value of 3.7369 and a 5% trimmed mean of 3.7432 (Appendix B, Table 21 – Normality Analysis – Descriptives). This indicated that the more extreme scores do not have a substantial influence on the mean. Further, the negative skewness and positive kurtosis indicates the score of the variable are clustered towards the right end side of the graph and that the distribution is rather peaked in the center with long tails and eventually some outliers. (Appendix B, table 21 and figure 4).
Parochialism had a mean value of 7.2385 and a 5% trimmed mean ranging of 7.2644 indicating there is no substantial impact from the extreme values on the mean. Furthermore, the negative skewness and the negative kurtosis indicate that the distribution of parochialism is clustered towards the right end and is not peaked in the center (appendix B, table 21 and figure 5). The results of the Kolmogorov-Smirnov statistics, which assess the normality of the distribution of scores, are summarized in appendix B, table 22. The values for the perceived pressure to perform are significant at the 0.000 level. This suggest the violation of the assumption of normality (Pallant, 2007). The normal probability plot for each of the variables show a reasonable straight line, suggesting normal distribution of the scores of perceived pressure to perform which are shown in appendix B, figure 6.
With regards to the levels of unethical behavior of employees, the results of the Kolmogorov-Smirnov statistics show that the value is significant at the 0.000 level, suggesting violation of the assumption of normality (appendix B, table 22). However, when analyzing the normal probability plot for the variables
we can see a relative straight line, signaling normal distribution of the scores as can be seen in appendix B, figure 7.
With regards to the moderator love of money, the results of the Kolmogorov-Smirnov statistics show that the values are significant at the 0.000 level. This suggests violation of the assumption of normality. The normal probability plots however show a straight line, suggesting normal distribution as shown in appendix B, table 23, and figure 8.
Regarding the moderator parochialism, the results of the Kolmogorov-Smirnov statistics show the values are significant at the 0.000 level, which suggest violation of the assumption of normality. The normal probability plots however show a straight line, which suggest normal distribution as can be seen in appendix B, table 23, figure 9.
Based on the values of perceived pressure to perform, presented in appendix B, table 21 and on the boxplots graphs (appendix B, figure 10), there are a total of 9 outliers – observations 25, 32, 100, 119, 138, 144, 152, 183, 187. These cases are considered outliers because it extends more than 1.5 times the box-lengths from the edge (Field, 2013). Looking at the boxplots for the levels of unethical behavior of employees (appendix B, figure 11) there are a total of 12 outliers – observations 69, 85, 114 121, 122, 123, 155, 157, 160, 161, 181, 199
Looking at the boxplots for the moderators love of money and parochialism, presented in appendix B, figures 12, 13 there were several outliers. For love of money observations were 14 outliers – observations 29, 32, 58, 69, 70, 82, 111, 116, 133, 147, 166, 195, 208 and 246. For parochialism 1 outliers was observed– observation 173.
Since the 5% trimmed mean value was relatively equal to the mean value for all of the variables in the analysis, the outliers and extreme values were not excluded from the data set. Therefore we proceed as if the data is normally distributed.
5.4 Correlations
Table 6 – correlation matrix variables, outlines the results from the bivariate correlations analysis of the variables. The most significant correlations have been marked with a single (p<0.05) or a double asterisk (p<0.01) (**). The relationships between the different variables were determined by using the Pearson product-moment correlation coefficient.
For the purpose of the analysis the categorical variables were transformed into dummy variables by assigning 1 and 0 for each of their levels. This resulted in the following variables that were used for the correlation analysis; unethical behavior of employees, perceived pressure to perform of employees, vignette 1, vignette 2, vignette 3, vignette 4, Love of money – rich, Love of money – motivator, Love of money – important, parochialism – horizontal collectivism, parochialism – vertical collectivism, age, gender – female, education – secondary school, education – vocational training, education – bachelor, education – master and education – other.
In the following paragraphs, the most important correlations with significant coefficients are discussed. First, the correlations between the dependent variable and the different vignette scenarios are discussed. Secondly, the most important correlations between the mediating variable and vignettes are outlined. Subsequently, the correlations between the control variables and
the dependent variables are outlined. Finally, some of the correlations between the moderating variables and the control variables are discussed.
When the unethical behavior of employees was tested in the different scenarios outlined in the vignettes there was a weak negative correlation in the scenario with no financial distress and no bonus system (Pearson correlation coefficient = -.159; N = 240; Sig. = 0.014). This would suggest if there is no financial distress in the firm and people are paid through a fixed salary they will behave less unethical.
Regarding the correlations between the moderating variables and the dependent variable, unethical behavior of employees, significant results were found for love of money (Pearson correlation coefficient = .155; N = 240; Sig. = 0.016), meaning people with high love of money are likely to behave more unethically. Interestingly, no correlations were found between love of money and perceived pressure to perform or between parochialism and the mediator or dependent variable.
Regarding the correlations between the mediator (perceived pressure to perform) and the different vignette scenarios significant results were found for the situation with both financial distress and bonus systems (Pearson correlation coefficient = .261; N = 240; Sig. = 0.000) and for the situation with neither financial distress nor bonus systems (Pearson correlation coefficient = -.440; N = 240; Sig. = 0.000). This suggest that perceived pressure to perform is higher when there is financial distress in the firm while bonus systems are used and lower when there is financial prosperity and people are being paid with a fixed salary.
Additionally, the analysis outlined positive correlations between several control variables and the dependent variable. Gender – male was shown to have a positive significant correlation with unethical behavior (Pearson correlation coefficient = .156; N = 240; Sig. = 0.015), indicating men are more likely to behave unethical than women. Furthermore a positive significant correlation was found between education – vocational training and unethical behavior of employees (Pearson correlation coefficient = .198; N = 240; Sig. = 0.002). This suggests people with vocational training as highest education are more likely to behave unethical than those with other educational training. Also, a positive significant correlation was found between love of money and parochialism (Pearson correlation coefficient = .179; N = 240; Sig. = 0.005) indicating people who score high on love of money also score high on parochialism.
Finally, significant correlations were found between control variables and moderators. Education – master and love of money was found significant (Pearson correlation coefficient love of money = .172; N = 240; Sig. = 0.008). This would suggest that people with education on the master level are more likely to score high on the personality trait love of money. In addition, significant correlations were fond between age and education – secondary training (Pearson correlation coefficient = -.174; N = 240; Sig. = 0.004) and between age and education – vocational training (Pearson correlation coefficient = .134; N = 240; Sig. = 0.027). These correlations suggest younger people in the sample are more likely to have completed secondary education, while the older people in the sample are more likely to have completed vocational training.
The full set of correlation is presented below in Table 6 – correlation matrix variables.
Table 6 - correlation matrix variables Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Unethical behavior of employees
-2. Perceived pressure of employees .300**
-3. Vignette 1 - Distress + bonus .078 .261**
-4. Vignette 2 - No distress + no bonus -.159* -.440** -.333**
-5. Vignette 3 - No distress + bonus -.024 .098 -.341** -.319**
-6. Vignette 4 - Distress + no bonus .100 .069 .348** -.326** -.333**
-7. Love of Money .155* .059 -.062 .079 .038 -.052
-8. Parochialism .123 .125 .077 -.028 -.136* .085 .179**
-9. Age .057 .012 .145* -.024 .006 -.129* -.020 .046
-10. Gender - female -.156* .061 -.086 .028 .022 .037 -.065 .031 -.136*
-11. Education - secondary school -.003 .041 -.046 .118 -.087 .017 -.107 -.054 -.174** .046
-12. Education - vocational training .198** .072 -.050 .052 .002 -.002 .054 .056 .134* -.142* -.086
-13. Education - bachelor -.059 -.003 -.067 -.099 .075 .091 -.108 .006 -.008 .135* -.473** -.256**
-14. Education - master -.027 -.076 .125 -.001 -.012 -.113 .172** .010 .083 -.132* -.218** -.118 -.648**
-15. Education - other -.022 .054 .108 -.036 -.037 -.038 .081 -.018 .021 -.045 -.024 -.013 -.072 -.033
-** Correlation significant at 0.01 level (2-tailed) * Correlation significant at 0.05 level (2-tailed)
5.5 Independent sample t-test
In order to test the first three hypotheses an independent sample t-test was conducted since the outcome (perceived pressure to perform) is continuous and there is one categorical predictor variable (bonus system or financial distress). First, dummy variables were created for the variables financial distress and bonus system; firms that experience financial distress (1), having no financial difficulties in the firm (0) and being paid through a bonus system (1) and having a fixed salary (0).
The first hypothesis that was tested is financial distress in the firm has a positive relationship with perceived pressure to perform of the employees. The results of the t-test can be found in table 7 and 8 below. Looking at Levene’s test for equal variances gives p < 0.05 so equal variances are not assumed. The p-value of 0.000 means there is a difference between the groups in terms of perceived pressure to perform when there is financial distress or no financial distress in the firm. This difference is .46737 on a scale of 1 – 5.
Table 7 – Group statistics financial distress on perceived pressure to perform
Financial distress N Mean Std. Deviation
Perceived pressure to perform No financial distress 116 3,4224 ,83331 Financial distress 124 3,8898 ,71343 Table 8 – Independent samples t-test financial distress on perceived pressure Levene’s test for equal variances T-test F Sig. Sig (2-tailed) Mean difference Perceived pressure to perform Equal variances assumed 4.422 .037 .000 -.46737 Equal variances not assumed .000 -.46737
