Amsterdam Business School
Master Thesis
Intrapersonal distributive justice related to the fairness perception
of incentives: an experimental study
Name: Jeffrey Voorend Student number: 10809031 Date: June 20th 2016 Word count: 14.891
MSc Accountancy & Control, Control
Statement of Originality
This document is written by student Jeffrey (Johannes Cornelis) Voorend 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
This study examines how the people’s fairness perceptions are influenced by the favorability of an intrapersonal comparison between the adjusted and unadjusted incentive. In the experiment, participants were randomly assigned to one of the two different conditions, where the favorability of the comparison was manipulated. In each condition several participants determined their fairness perception of a hypothetical situation. The manipulation is related to the favorability of the comparison between the adjusted incentive and the unadjusted incentive. The experiment resulted in a significantly higher fairness perception of the adjusted incentive when the incentive was related to a favorable intrapersonal comparison, than when the incentive was related to an unfavorable comparison. It should be noted that this difference is based on self-interest, since both experiment conditions were equally fair. A third (neutral) experiment condition without any subjective adjustment was perceived as the fairest according to the participants. This high fairness perception is possibly related to the controllability of the performance measure. This suggest that a subjective adjustment of an uncontrollable performance outcome restores the fairness perception of the outcome. However, it will not reach the fairness perception of a performance outcome that is considered as controllable. This could be an interesting topic for future research.
Index of tables and figures
Figure 1: Libby boxes…...……….11
Table 1: Descriptive statistics………..……….22
Table 2: Inter-item correlation matrix………..………24
Table 3: Independent sample t-test of IFP’s between regular experiment groups…...…..25 Figure 2: Mean IFP per experiment condition...…...………26 Table 4: Statistically significant differences between experiment groups (LSD method)...27
Contents
Abstract ... 3
Index of tables and figures ... 4
1 Introduction ... 7
2 Literature ... 11
2.1 Prior literature ... 11
2.1.1 Fairness perception ... 11
2.1.2 Intrapersonal distributive justice principle ... 13
2.2 Hypothesis Development ... 15 3 Methodology ... 17 3.1 Participants ... 17 3.2 Experimental design ... 18 3.3 Procedure ... 20 3.4 Validity threats... 22 4 Results... 24 4.1 Statistical analysis ... 24 4.1.1 Reliability analysis ... 24 4.1.2 Hypothesis test ... 24 4.1.3 Supplemental analysis ... 26 4.1.4 Control variables ... 27
4.2 Implications of the results ... 28
4.2.1 Hypothesis test ... 28
4.2.2 Supplemental analysis ... 29
4.3 Limitations of the results ... 30
5 Summary and conclusion ... 32
7 Appendices ... 38
7.1 Unfavorable comparison with intrapersonal referent (English) ... 38
7.2 Favorable comparison with intrapersonal referent (English) ... 41
7.3 Control condition (English) ... 44
7.4 Unfavorable comparison with intrapersonal referent (Dutch) ... 47
7.5 Favorable comparison with intrapersonal referent (Dutch) ... 50
7.6 Control condition (Dutch) ... 53
1 Introduction
In this research, I examine to what extent the favorability of an intrapersonal comparison affects the people’s fairness perceptions.
Fairness perceptions of individuals, such as employees, are often used as an outcome variable in economics (Paul, 2006). A fairness perception contains behavioral implications and has subjective characteristics. The human factor such as personal values and beliefs may play an important role in determining the fairness perception (Paul, 2006). A feeling of justice is important when compensation decisions, such as pay, pay raises and bonuses have to be determined (Jawahar & Stone 2011). The fairness perception may not represent the actual fairness in some cases, but this perception has, however, many consequences. A low fairness perception can result in a lack of motivation, reduced effort and worse attitude of employees (Bol & Smith, 2011; Simons & Roberson, 2003; Sweeney, 1990). On the other hand, an employee’s performance can improve when their fairness perception is higher (Bol et al. 2015; Hartmann & Slapničar, 2012; Simons & Roberson, 2003). On top of that, fairness perceptions can have consequences for the level of productivity. Individuals are more inclined to quit and search for jobs that reflect the true value of their productivity in a better way (Paul, 2006). Therefore the fairness perception of pay is important for organizations and employees. It can influence the work motivation of employees, which changes the employee’s performance as a result of that (Bregn, 2013). In times of growing employee empowerment, the fairness perception of the employee becomes even more important (Sanchez & Cralle, 2012).
The ambition of this study is to examine to what extent an intrapersonal comparison, regarding a performance evaluation and incentive, will be taken into account by individuals concerning their fairness perception. Since I am not aware of any other research with this specific focus and setting, this is also the key contribution of the paper. In this study, the equity theory and the self-interest bias are used to explain the aforementioned relationship.
This study conducts an experiment with random people as participants. The experiment corresponds with a case setting, where the participant has to imagine he/she is a car dealer manager, who receives an incentive according to his/her performance. This experiment analyses the people’s fairness perceptions of their incentive in two different conditions, where the influence of the uncontrollable event to the related ex-post subjective adjustments of performance outcomes is manipulated. This performance outcome is the result of a performance measure and leads to the incentive. According to the case setting, the participant has to determine his/her fairness perception of the incentive by answering three questions on a seven-point Likert scale. The mean value of the answers corresponds with the individual’s fairness
perception. The experiment employs a 1 × 2 + 1 between-subjects design. On top of the two regular conditions, there is a third (control) condition. This condition only contains information about an objectively measured performance outcome, which represents the contribution of the individual to a reasonable degree. In this condition no uncontrollable event and adjusted performance outcome is applied.
Each participant was unknowingly and randomly assigned to one of the three conditions, which has resulted in answers that are independent from each other. The case states what the car dealer manager was receiving when the performance outcome and the related incentive was purely objective and calculation-based. Furthermore, the participants were aware of the uncontrollable event and of the related subjective adjustment, which was based on legitimate reasons.
The participants had to imagine that they were the manager of a car dealership, who had the aim to sell a new innovative product. According to the sales quantity, the manager would receive an incentive at the end of the year to motivate him/her during the year. The intention of the incentive was that its size has to represent the manager’s input to the company. Nevertheless, the performance outcome of the manager became noisy due to an unforeseen uncontrollable event. That event resulted in a subjective adjustment of the performance outcome and the related incentive. The manipulation is related to whether the uncontrollable event affects the performance outcome and incentive in a positive or negative way. Due to the uncontrollable event the objective calculation-based reward-to-investment-ratio (RI-ratio) could increase or decrease. The subjective adjustment compensates noise by adjusting the incentive upwards or downwards. The final performance outcome and the related incentive is of the same magnitude and is equally fair - the RI-ratio is the same - in every condition. Despite that I expect that the fairness perceptions differs when the favorability of the comparison with the intrapersonal referents also differs. In other words, a higher fairness perception is expected when the comparison with the intrapersonal referent is favorable for the individual. This expectation is based on literature, which will be discussed in section 2.
This experiment examines whether and how the comparison with an intrapersonal referent can influence the fairness perception of individuals. In this study the intrapersonal referent is created by subjectively adjusted incentives, which results in a comparison between an adjusted measure and unadjusted measure. This comparison is the key element of the
unadjusted measure/incentive. A subjective adjustments can be used when the objective performance measure gives an outcome that does not represent the actual input. Then the measure can be adjusted after the outcome is known (Merchant & Stede, 2012). The use of the so-called ´ex-post subjective adjustments´ occurs regularly. The already mentioned importance of the fairness perception in combination with the common used subjectivity makes this study important and relevant.
The main focus of this study is related to the trade-off between the equity theory and the self-interest bias. The relation of the intrapersonal distributive justice and the people’s fairness perceptions corresponds with the self-interest theory. If a favorable comparison with a referent always leads to a higher fairness perception, then this perception is based on self-interest (Messick & Sentis, 1979). If a favorable comparison is not significantly taken into account, it can be suggested that the equity theory outweighs the self-interest bias (Adams, 1965). The equity theory concentrates on the RI-ratio, which is the comparison of the subordinate’s reward and contribution to the company (Adams, 1965). This ratio has to be close to one to make the situation the fairest (Adams, 1965). This study examines to what extent individuals may attach value to favorable intrapersonal referents when they determine their fairness perception. In other words, to what extent do individuals take into account their unadjusted incentive when they determine the fairness of their adjusted incentive?
The results of this experiment can be valuable in a working environment where the use of ex-post subjective adjustments occurs regularly. Subjectivity can improve incentive compensations by offsetting uncontrollable factors (noise) (Bellavance et al. 2013). According to Ittner et al. (2003) subjective adjustments can improve incentive contracting by exploiting non-contractible information that might be ignored by purely calculation-based reward systems. Furthermore, subjectivity can reduce dysfunctional behaviour caused by incomplete objective performance measures (Ittner et al. 2003). A subjective adjustment is often caused by an uncontrollable event, which makes the performance outcome noisy. It is important to take the fairness opinion of the employee into account when subjective adjustments are used in combination with objective measured performance outcomes. Although fairness perceptions may be biased through self-interest, it is still an important factor which can significantly influence people’s motivations, attitudes, job satisfaction and other emotions (Bol & Smith, 2011; Simons & Roberson, 2003; Sweeney, 1990).
The average fairness perception outcome of each of the three experiment conditions results in an answer to the following research question: To what extent do individuals consider a favorable intrapersonal comparison when they determine their fairness perception? The purpose
of this research is to answer the question by examining to what extent a favorable intrapersonal comparison with the unadjusted incentive is considered when individuals determine the fairness perception of their adjusted incentive. When a favorable intrapersonal comparison leads to a higher fairness perception, the self-interest bias is present.
The findings support the expectation about the favorability of the intrapersonal comparison related to the people’s fairness perceptions. An independent sample t-test revealed a statistically significant difference in the fairness perceptions between the experiment groups with favorable and unfavorable comparisons. The present study provides experimental evidence that the self-interest bias occurs when an intrapersonal comparison regarding an incentive is made. In addition to that, a third condition without any subjective adjustments was perceived as fairer than the regular conditions with subjective adjustments. This is probably related to the performance measure, which seems to be more controllable in the control condition. An unadjusted controllable outcome may be seen as fairer than a subjectively adjusted uncontrollable outcome. The last mentioned suggestion could be potential material for future research.
Section 2 describes the prior literature and hypothesis development of this study. Section 3 describes the methodology and the research design. Section 4 illustrates the results and finally section 5 concludes and summarizes the paper.
2 Literature
The first part of this section describes the prior literature about the two research constructs that are linked in this study. These constructs are the intrapersonal distributive justice principle and the people’s fairness perceptions. The literature review results in the hypothesis, which combines the two mentioned constructs. The equity theory and the self-interest bias will help the hypothesis and the related research question getting answered.
2.1 Prior literature
The first part deals with the dependent variable of this research, which is the individual’s fairness perception (IFP). The IFP is related to the equity theory of Adams (1965) and can possibly contain a self-interest bias. It is crucial in this study to get a good view of the theories in combination with the research areas, so this theory and bias will be discussed in the first part.
The intrapersonal distributive justice principle could already be interpreted as an element of the people’s fairness perceptions. However, not many research is done on the actual influence of this principle to people’s fairness perceptions. Therefore the independent variable relates to the aforementioned principle. To be more precise, the independent variable is the favorability of the intrapersonal comparison. This will be discussed in paragraph 2.1.2.
In this study the focus especially lays on the attached value that people could have to a favorable intrapersonal comparison when they determine their fairness perception. In the figure below the combination of the two research areas with the related measures is shown.
Figure 1: Libby boxes
2.1.1 Fairness perception
becomes more and more dependent of their employees due to increasing dynamic environments with, in some cases, high technical developments nowadays (Mendes & Machado (2015). The employees’ performance can significantly influence the performance of the company they are working for. Moreover, the employees’ fairness perceptions are related to their performance (Bol et al. 2015; Hartmann & Slapničar, 2012; Simons & Roberson, 2003). Thus, the employees’ perceived fairness becomes more important for organizations.
Organizational justice can be divided in distributive justice and procedural justice. These justices are investigated by Bos et al (2001) who refer to Greenberg (1986) as well as Sheppard and Lewicki (1987). They mention that people make a distinction between issues related to the distributive justice and procedural justice when they determine their fairness perception. The procedural justice determines the overall fairness of the selection process, while distributive justice represents the overall fairness of the selection outcome (Patterson et al. 2011). Research has proven that there is an influence of procedural justice on fairness perceptions of performance evaluations and incentives (Greenberg, 1986; McFarlin & Sweeney, 1992). In this study, procedural justice will be held constant and will not be mentioned further.
Distributive justice is equal to a person’s view whether his outcome or incentive is fair by comparing the outcome to a referent (Huffman & Cain, 2001). The main focuses of distributive justice are on the equity norm, the equality norm and the need norm (Murphy-Berman et al. 2012). The equality norm means that people should be treated as equals, for example when incentives are allocated among them. However, this norm is not relevant in this study, since there is no interpersonal referent in this experiment.
The equity norm, which relates to the equity theory of Adams (1965), is more relevant here. The equity theory refers to the RI-ratio, which can be interpreted as the proportion of what an individual receives as incentive (reward) in exchange for the performance he delivers (investment) (Taris et al. 2002). The equity theory assumes that an inequitable outcome can result in a perception of unfairness, whereas an equitable outcome leads to higher fairness perceptions (Loseman et al. 2009; Miles et al. 2015; Sweeney, 1990). This theory contradicts the self-interest bias, which states that the fairness perception is the highest when the output is in line with people’s needs, regardless of their input (Ubeda, 2014).
As already mentioned, Murphy-Berman et al. (2012) divide the distributive justice principle in the equity and equality theory and in the need norm. The equity and equality theories
expressed as determining the IFP by weighting different motives in a way that it is in the individual’s own interest (Cappelen et al. 2005; Cole & Flint 2004; Ubeda 2014). Both descriptions suggest that the fairness perception is higher when the needs of the people (high incentives) are met. First, the self-interest bias was only proposed by psychologists, now it is also accepted in the economical science (Ubeda, 2014).
Moore and Loewenstein (2004) mention that the self-interest bias can occur automatically and involuntarily, but also process controlled and voluntarily. Thus, the people’s judgments are likely to be influenced by self-interest subconsciously. Many studies find that distributive justice is associated with the self-interest bias. Verboon and van Dijke (2007) conclude that self-interest considerations determine why people care about distributive justice. Furthermore, Ubeda (2014) shows that an individual perceived his/her incentive as fair, when the incentive is, in comparison with others, favorable and in line with his/her own interest. Messick and Sentis (1979) show that individuals consider it as fair when they are paid more than the other individual, even when they both work the same amount of hours. In other words, the distributive justice principle can contain an egocentric perspective (Rodriguez-Lara & Moreno-Garrido, 2012).
The aforementioned studies suggest that an interpersonal referent may influence the IFP’s and that the self-interest bias subconsciously may play a role in that influence. In conclusion, a comparison with a referent may include a self-interest bias. The studies mentioned so far, concentrate on the interpersonal aspects whereas the present study concentrates on the intrapersonal aspect. However, it suggests that the self-interest bias may also occur on the intrapersonal aspect.
2.1.2 Intrapersonal distributive justice principle
Referents of distributive justice may be interpersonal or intrapersonal. The comparison with interpersonal referents is the comparison of people’s performance outcome and input (time, effort etc.) to that of other people. Research about distributive justice mostly concentrates on this variant. And components of distributive justice, such as the equity and equality theory, initially focus on the comparison with other persons too (Adams, 1965; Murphy-Berman et al. 2012; Patterson et al., 2011). However, in this study the relevant item is the intrapersonal distributive justice principle, which means that a person uses a referent related to himself to determine his fairness perception (Huffman & Cain 2001). Folger (1986) suggests in his referent cognition theory that intrapersonal referents may arise from past experiences or other sources, such as comparing an outcome to another outcome that might have occurred when a different measure had been used. Huffman and Cain (2001) mentioned that an unadjusted measure can be
used to create an intrapersonal referent. In this study, the last mentioned intrapersonal referent is relevant.
Little research has been done on intrapersonal distributive justice, but one of the most relevant researches is that of Huffman and Cain (2001). They find that an intrapersonal referent affects the satisfaction with the performance outcome. In their experiment individuals were dissatisfied with their outcome if they believed that the performance outcome would be better under a different situation. Yet it is important to distinguish satisfaction with and fairness perception of a performance outcome. Because an individual can be satisfied with an outcome, but at the same time thinks that it is unfair or vice versa (Paul, 2006). However, it suggests that the intrapersonal distributive justice principle is also a determining factor in measuring people’s fairness perceptions.
Furthermore, O’Malley (1983) finds that fairness perceptions which are already grounded on interpersonal referents, could also be influenced by intrapersonal referents. In his experiment, every participant judges the fairness of his bonus by taking another person’s bonus into account. When the comparison with that particular interpersonal referent is not perceived as fair, the favorable comparison with an intrapersonal standard (the expected bonus) compensates the feeling of injustice. However, in contrast to O’Malley’s study, the present research excludes the interpersonal referent. Moreover, instead of an expected bonus, an unadjusted measure will be used as intrapersonal referent. A comparison between the adjusted and unadjusted measure in response to an uncontrollable event can be more relevant in case of uncertain working environments. Besides, O’Malley’s experiment contains no further specifications about the real fairness as mentioned in my study. In my experiment the participants know that the subjective adjustment is based on legitimate reasons, so it is assumed that the adjusted measure in both experiment conditions are actually fair. To what extent the people’s perceived fairness is different, is the question I added in the present study.
The aforementioned studies suggest that the IFP is higher when the intrapersonal comparison is in the individual’s advantage, than when the comparison is in the individual’s disadvantage. This could even be the case when the adjusted performance outcomes and related incentives are actually equally fair. So in the present study the adjusted performance outcome may be equitable regarding to the RI-ratio of Adams’ equity theory. But the self-interest bias may cause a higher IFP due to a favorable intrapersonal comparison.
2.2 Hypothesis Development
The situation in this study’s hypothesis corresponds with the trade-off between the self-interest bias and actual justice (equity theory) (Sauermann & Kaiser, 2010). Actual fairness with an RI-ratio close to one may not always be seen as fair according to the individual. He may perceive an increased RI-ratio as fairer, because that is favorable according to the need norm of Schwinger (1980). After all, an increased RI-ratio maximizes their personal welfare. Comparing the adjusted performance outcome with the intrapersonal referent (the unadjusted performance outcome) is described as the intrapersonal distributive justice principle in this study. This principle shows whether people, when they determine their fairness perception, take the favorability of the comparison with a referent, related to the person himself, into account. Whether and to what extent the intrapersonal comparison leads to a self-interest biased fairness perception, is what this experiment examines.
Participants of the experiment will read a situation where they have to imagine that they face an upward adjusted incentive or a downward adjusted incentive. This incentive corresponds with the performance outcome. In both situations the final adjusted performance outcomes and incentives represent the contribution of the participant (as a car dealership manager) to the same extent. The actual fairness of both incentives are the same too. However, due to the self-serving bias the individuals may perceive the fairness of the same incentive different, only because one incentive is subjectively adjusted upwards whereas the other incentive is subjectively adjusted downwards.
Factors like a cognitive distortion (Bol & Smith, 2011) could have an effect on subjective adjustments, which makes that the adjustment is unjustified on beforehand. However, in the present experiment it is assumed that all of the subjective adjustments are based on legitimate reasons. Thus, the subjective adjustments are equally fair. However, the IFP of the participants may still differ, even when the adjustments are legitimate.
According to the equity theory (Adams, 1965), the average fairness perceptions of the two experiment conditions have to be equal. But based on prior literature it is more likely to predict that the self-interest bias occurs, be it subconsciously. When hypothesis one is supported, the incentives that are subjectively adjusted upwards result in a higher average of the IFP’s, than the incentives that are subjectively adjusted downwards. Only the favorability of the comparison with the intrapersonal referent could be the cause of a difference in IFP’s, because all other variables remain constant in the experiment. The hypothesis of this research is:
H1: The average fairness perception of an adjusted incentive will be higher when the intrapersonal comparison with the unadjusted incentive is favorable, than when the intrapersonal comparison is unfavorable.
3 Methodology
This section discusses the participants and the experimental design of this study. Furthermore, the procedure of generating the results and the validity threats of the experiment will be discussed.
3.1 Participants
To join the experiment as a valid participant, the individual had to be part of the working age population. Joining the experiment could be done by completing a three-item questionnaire about incentives in a car dealership. 88 people (n=88) participated in the experiment with the average age of 36 years. The main variable in this experiment is the people’s feeling of justice about a certain situation regarding the adjustments of bonuses. Since the experiment’s case setting is only relevant for business people, the participants had to be part of the working age population, while participating in the experiment. Because all participants turned out to be Dutch, the sample of the present study is focused on the Dutch working age population.
Inexperienced students as well as experienced workers could participate in the experiment. Students may have acquired more theoretical knowledge to make a rational consideration when they determine their fairness perception. But in contrast to students, people with more work experience could use their experience to base their fairness perceptions on. Situations regarding the perceived fairness of incentives are relevant in every business layer. Whether someone is a top manager or an employee, incentives could influence the work motivation and productivity of both individuals (Bregn, 2013; Paul, 2006). Therefore people out of the whole working age population could participate in this experiment.
People had to fill in demographic information first in order to observe whether this research obtained an appropriate sample for the experiment. Demographic information such as gender, age and education have to represent the average working age population in a reasonable way to strengthen the validity of the results. Furthermore, this information was assessed to control for a potential impact.
The participants of the experiment had an average age of 36 years, which is not the same as the average age of the Dutch working age population (41.9 years in 2014) (CBS Statline, 2016). However, it reflects the average age of the Dutch working age population in a reasonable way. Furthermore, in this experiment the ratio of participants with a higher level education (HBO or higher) versus people with a middle/lower level education (MBO or lower) was 35% to 65%. In 2015 this ratio in the Dutch working age population was exactly the same (CBS, 2015). Also the
participant’s gender ratio of 59% male and 41% female gives an reasonably accurate picture of the gender ratio of the Dutch working age population (54% male / 46% female in 2015) (CBS, 2015).
The experiment exists of three slightly different questionnaires. Several participants filled in one of these variants, which put the participants into three different groups. For the experiment’s results it is important that the demographic information (gender, age, education, etc.) of those three groups is not too much deviated from the average information of all the 88 participants. If the demographic information of the three experiment groups is too much deviated from each other, the results of the experiment could be biased (Vella, 1998). To prevent this bias, the participants are randomly assigned to an experiment condition. This increases the internal validity of the experiment.
A total of 88 people participated in the experiment. 83 people filled in the questionnaire completely, who are equally divided in three groups of 28, 28 and 27 participants. To control for other personal characteristics, such as capability to empathize in a case setting, I relied on random allocation of participants into the three experiment groups. This random allocation was done by the survey program.
3.2 Experimental design
The experiment employs a 1 × 2 + 1 between-subjects design. In this design I manipulated the intrapersonal comparison between the adjusted and unadjusted measure at two levels (unfavorable versus favorable comparison). The manipulation was caused by an upward and downward subjective adjustment of the performance evaluation and incentive. Without the participant’s awareness, each one of them is randomly assigned to one of the two conditions or the control condition.
The formats are composed as equally as possible. In every condition the participants were told to role-play that they were a car dealership manager (subordinate) whose performance was evaluated and rewarded by the regional manager (supervisor). In both of the regular experiment conditions, participants received information about an agreed incentive, which was 2% of the new car series’ realized profit. A subjective adjustment of the bonuses has occurred in both of the conditions too, either upwards or downwards. In addition, the only difference between the experiment conditions is the manipulation of the comparison with the intrapersonal
event has occurred, which has a positive effect on the manager’s incentive. Therefore the incentive increases from €120.000 to €160.000. The uncontrollable event is described as an unexpected decrease of near competition, which results in an increase of the sales amounts of new cars. The increased sales amounts of the new cars, in turn, results in an increased incentive of the car dealer manager. In response to that the regional manager decides to adjust the incentive downwards to undo the effects of the uncontrollable event. The regional manager’s decision results in a negative comparison of the actual incentive with the unadjusted incentive. In the favorable comparison condition the exact opposite of the aforementioned situation occurs. An uncontrollable event results in a decrease of the manager’s incentive from €120.000 to €80.000. Therefore the regional manager decides to adjust the car dealer manager’s incentive upwards, which results in a favorable comparison with the unadjusted incentive. In every experiment condition the adjusted incentive is the same amount (€120.000). There is no knowledge of the participants about the existence of other conditions. So the answers of each experiment condition are independent from each other.
This aforementioned information regarding the uncontrollable event and the incentive adjustment was all known by the participants, who had to determine their IFP based on the given information. Both conditions are equally fair according to Adams’ equity theory. The adjusted incentive in every condition is the same size and the case emphasize that the adjustments are based on legitimate reasons. Despite the equal fairness of the conditions, the average of the IFP’s in these conditions could be different, merely because of the different intrapersonal comparison.
Next to the two regular experiment conditions there is the control condition, which is, also in terms of actual fairness, comparable to the regular conditions. However, in the control condition no uncontrollable event and incentive adjustment are present. In the control condition the manager’s incentive is not adjusted, because of the regional manager’s uncertainty about whether the incentive is too much or too less compared to the manager’s input. The incentive in this condition remains €120.000.
The average IFP’s of the two regular experiment conditions could be compared to each other to answer the research question, but due to the presence of the control condition the results analysis of this study is extended. Now the two regular variants could also be compared with the neutral condition in order to determine whether an intrapersonal comparison affects the fairness perception in general.
After reading all the case information, participants were asked to determine their IFP’s of the hypothetical case setting by answering three questions on a seven-point Likert scale. This
scale makes the dependent variable measurable. The mean value of the three answers represents the IFP’s of one participant. The first question relates to the justification of the decision to adjust the incentive. The second one is about the comparison between the actual incentive and the manager’s contribution to the company. The third and last statement is about the overall fairness of the adjusted incentive. All of the three questions corresponds with the feeling of justice about the incentive of the participant. That is confirmed by the Cronbach’s alpha of the standardized items. The Cronbach’s alpha is 0.841, which can be interpreted as above sufficient. Therefore no question has to be removed from the analysis. The averaged and standardized value of all the answers per experiment condition is called the average IFP, which is the dependent variable of this study.
The essence of this experiment is to test whether and to what extent the IFP of the participants is affected by the intrapersonal comparison between the adjusted and unadjusted incentive. If the study’s hypothesis turns out to be supported, the average IFP of the group with the favorable intrapersonal comparison would be significantly higher than the average IFP of the group with the unfavorable intrapersonal comparison. The primary aim of this quantitative research is to investigate whether the averages of the IFP’s between the two regular experiments groups differ significantly. To take care of well-interpreted results regarding the hypothesis, the two-tailed independent sample t-tests will be used.
3.3 Procedure
In order to get an appropriate sample, which represents the Dutch working age population, the three experiment conditions of the experiment were shared online. I shared the questionnaires with various connections, such as family, friends and colleagues. The questionnaires corresponds with a URL, which I shared on, for example, Facebook. By clicking on the URL, the survey program SurveyMonkey randomly assigned each participant to one of the three conditions. To prevent the experiment from manipulation, the existence of different experiment conditions were not revealed to the participants. The results of the experiment were administered automatically by the survey program, which ensures there were no administration errors. When enough participants responded, the results were copied in the SPSS-program.
The questionnaires comprised three parts. The first part of the questionnaire was a brief introduction of what will be expected from the participant. The main part existed out of the
asked. To prevent this research from manipulation, participants were not informed about the contents of the questionnaire until they opened the URL. Each participant could only fill in one questionnaire. This was regulated by the survey program, which allowed to fill in only one questionnaire per IP address. Overall, the participants spent approximately 5 minutes completing the questionnaire.
Before the questionnaires were shared among the participants, the case settings were pre-tested by three objective persons in order to investigate whether they would understand the questionnaire. First, each person tested individually one case setting, after which the persons looked into all cases as a group. After a few adaptations, the questionnaires could be shared. Next to the English versions of the questionnaires, Dutch versions are created so that non-native English speakers could also understand the case setting. Both English and Dutch versions are presented in the appendices.
In the case setting, the car dealer manager, ergo participant was charged with the task of selling as many new electric cars as possible. The participants were told that the manager’s performance was evaluated and rewarded following the new cars’ realized profit. The evaluation and rewarding was done by the regional manager. An uncontrollable event during the year affected the manager’s compensation. This resulted in an inaccurate reflection of the manager’s contribution to the company and his incentive. Through that inaccurate reflection the regional manager decided to depart from the compensation agreement and adjust the car dealer manager’s incentive. The case settings emphasize that the incentive adjustment was based on legitimate reasons, which makes that the RI-ratio, hence the actual fairness, is equal in all of the experiment conditions.
Through the given information in the case setting, the comparison between the adjusted measure and the unadjusted measure could be considered by participants when they determined their IFP. The (un)favorable comparison with the intrapersonal referent corresponds with the independent variable of the present study. The favorability of the comparison makes the independent variable measurable.
The data gathering turned out to be quite straightforward. The questionnaire is short and to-the-point, which resulted in that participants did not run out of motivation to complete the questionnaire. Therefore there is also a high probability that participants made well-considered decisions, while answering the questions of the case setting. Despite that, not every questionnaire was filled in completely. Five out of the 88 participants forgot to answer the demographic questions at the last part of the questionnaire. The demographic information is important in order to observe whether the participants properly represents the Dutch working age population.
Therefore I eliminated the five incomplete questionnaires from the analysis. After eliminating them, 83 answers remained valid.
The IFP of every participant is determined by taking the mean value of the three questionnaire items. The average of all the IFP’s are separated per experiment group. This is called the average IFP (per group). A higher IFP indicates a higher feeling of justice regarding the given case setting. The descriptive statistics of the dependent variable are presented in table 1.
Table 1: Descriptive statistics
3.4 Validity threats
According to Messick (1995), the validity of a study relates to how the score meaning, of the experiment in this case, is hold across persons or populations. There are many factors that could interfere with the assessment of the data, which leads to a lot of possible validity threats (Downing & Haladyna, 2004). Out of a number of sources the most important validity threats are selected, after which the proposed measures are mentioned in this paragraph (Downing & Haladyna, 2004; Maas, 2011; Saal et al. 1980). Next to the large sample of 88 participants, additional precautions were taken to verify the robustness of the results:
To prevent the selection bias that is mentioned by Maas (2011), participants are randomly selected. The fact that the demographic information of the participants represents the working age population in a reasonable way, confirms the absence of this bias.
Furthermore, the formats of every experiment condition were composed as equally as possible. All cases were structured in a way that a well-considered fairness opinion should be made by the participants whereby the intrapersonal referent is clearly explained. The only difference in the experiment conditions was the manipulation of the favorability regarding the intrapersonal comparison. Every case setting was of the same size and the scenarios were equally interesting, which ensures that all experiment conditions were equally likely to finish. Thus, also
In the brief introduction of the questionnaire it is mentioned that the answers and personal information will not be showed to other people. Therefore the participants had no motives to bias their answers consciously. Whether the participants bias their answers subconsciously out of self-interest is more or less the research question of this study and will be discussed in section 4.
Next to the aforementioned validity threats, there is the central tendency error. This could be described as the tendency of the participant to rate everything around the midpoint of the (Likert) scale (Saal et al. 1980). This error could lead to middle-of-the-road results, which threats the study’s validity, because there may not be found any significant differences between the mean IFP’s of the experiment groups. In the result analysis of this study it can be concluded whether the central tendency error actually deteriorates the research validity.
Finally, the leniency or severity effect and the halo effect are regularly related to studies where people have to judge other people (Saal et al, 1980; Wolf, 2015). Since this is not the case in this experiment, these validity threats can be excluded from the present study.
4 Results
This section analyses the data from the questionnaires to examine whether the hypothesis could be rejected or not. This section ends with the implications and limitations of the results.
4.1 Statistical analysis
The statistical analysis discusses the reliability of the analysis, after which the hypothesis will be tested. Furthermore, this paragraph discusses the supplemental analysis and the participants’ demographic information.
4.1.1 Reliability analysis
Before the analysis of the results takes place, the reliability of the results has to be determined. Cronbach’s alpha is used to ensure whether the results are reliable. The Cronbach’s alpha estimates the internal consistency of each concept of the experiment scores. The aim of the questionnaire items of this experiment is to measure the same construct. In this questionnaire there is only one construct to measure, which is the IFP. This independent variable is measured by three questionnaire items, none of which has to be reversed.
The Cronbach’s alpha of the questionnaire items is 0.841, which is well above the threshold of 0.7. According to Field (2009), a Cronbach’s alpha between 0.8 and 0.9 is great, indicating that the outcomes of all three questionnaire items represents the IFP in a more than sufficient way. Table 2 shows that every correlation of the three questionnaire items is significant to the other. The Cronbach’s alpha could be higher (0.865) when the second question, regarding the comparison of the manager’s incentive and contribution to the company, was deleted from the analysis. However, the present Cronbach’s alpha is already above sufficient and the possible increase of the alpha is not so dramatic that the three questionnaire items have to be reduced to two.
perception of an adjusted incentive. The two regular experiment variants clearly explained that the participant, as a car dealer manager, had to deal with an uncontrollable event which led to the justifiable decision of the regional manager to adjust the car dealer manager’s incentive upwards or downwards. The related hypothesis predicts that the upward adjusted incentive results in a higher average of the IFP’s than the equally fair downward adjusted incentive.
The input and output, which is the manager’s contribution to the company and the incentive, is the same in both conditions. In the case is mentioned that the effect of the uncontrollable event was not the result of a change in the manager’s performance. Despite that, a difference is expected regarding the perceived fairness, which is caused by self-interest.
An independent sample t-test is conducted to determine whether there is a statistically significant difference between the mean IFP’s of the two regular experiment groups. Table 1 in section 3 and figure 2 show that the mean IFP’s of the experiment groups with the favorable and unfavorable comparisons were respectively 4.01 and 3.04. The t-test compares the means of the same dependent variable of two unrelated groups: the IFP. Because the Levene’s test for equal variances provides a p-value of 0.380 (p > 0.05), the equal variances could be assumed. Table 3 shows that the corresponding two-tailed significance is 0.009 (t = 2.700; p < 0.01), which makes that the test revealed a statistically significant difference between the two regular experiment groups in relation to the IFP.
To sum up, H1 is supported, demonstrating that the IFP of the case regarding the favorable comparison is significantly higher than the IFP of the case regarding the unfavorable comparison. This implies that the null hypothesis, which assumes that there is no significant difference in IFP between both experiment groups, has to be rejected. The theory and related expectations, mentioned in section 2, are confirmed in the t-test. It suggests that, regardless of the equal RI-ratio and the legitimate reasons of the adjusted incentive, people will significantly take into account the intrapersonal comparison between the adjusted and unadjusted incentive, when they determine their IFP. Thus, the self-interest bias is present in this experiment.
Figure 2: Mean IFP per experiment condition
4.1.3 Supplemental analysis
In the previous paragraph the hypothesis of this study is tested. The hypothesis is merely related to the two regular experiment conditions. Additionally, there is the existence of the control condition, in which the participants were also asked about their IFP. A subjective adjustment of an incentive is not relevant in this condition. In order to gain a better understanding about the IFP’s of a subjectively adjusted incentive compared to the IFP’s of no adjusted incentive at all, this neutral condition is created. The mean IFP of the control condition is 4.86 on a seven-point Likert scale, whereas the IFP’s of both regular experiment conditions are lower (see figure 2). Whether this difference is statistically significant, is the main question in this paragraph.
Because of the addition of a third condition, an independent sample t-test is not relevant anymore. To test for differences between more than two groups, a one-way analysis of variance (ANOVA) is the best option. Out of the one-way ANOVA it can be concluded that the mean IFP of at least one experiment group significantly differ from another group (F = 13.751; p < 0.01). However, this is already concluded is the previous paragraph, which showed that the group with the favorable comparison differs significantly from the group with the unfavorable comparison.
A further analysis is necessary to conclude whether there are more than one significant differences between the three groups. Therefore a Post Hoc analysis with Fisher´s Least
experiment groups (p < 0.05). The significant difference between the favorable and unfavorable comparison is already mentioned. However, table 4 shows that the IFP of the control condition is significantly different from the IFP’s of both regular conditions. The statistical differences between all of the three experiment conditions mean that the possible presence of the central tendency error does not have consequences for the study’s results.
Table 4: Statistically significant differences between experiment groups (LSD method)
The mean value of the IFP of the control condition is higher than the IFP’s of the two regular conditions. This indicates that subjective adjustments result in a decrease of the mean IFP. Even an upward subjective adjustment of an incentive is perceived as less fair than no subjective adjustment at all.
4.1.4 Control variables
The previous paragraphs discussed possible significant differences between the three experiment groups. Additionally, control variables, such as gender and age, could influence the dependent variable. The personal information of every person that participated, is collected. At first, this was in order to observe whether the participants represent the working age population. But furthermore, this information is necessary to control for a potential impact in IFP. It will be included in the result analysis, when it appears that there is a statistically significant difference in IFP’s between people with, for example, different levels of education.
The control variables that has been asked for are gender, age, the nationality, years of work experience and the highest level of education. As already mentioned, the nationality of the participants all turned out to be Dutch, so the nationality is excluded from further analysis.
By means of independent sample t-tests, I examined whether one of the remaining control variables influence the IFP significantly. The tests revealed no statistically significant results for any of the control variables (p > 0.05).
The last mentioned t-tests were taken over the whole sample in general. However, for each of the control variables a difference in the IFP’s could also be present within only one experiment group. Whether this is the case, is analysed by a factorial ANOVA. Out of that analysis, there can be concluded that there are no significant effects between any of the control variables and IFP’s within the experiment groups (p > 0.05). In conclusion, there is no control variable or any specific combination of a control variable and an experiment group that
significantly affects the IFP. The experiment group itself is the only variable that influences the IFP significantly.
4.2 Implications of the results
The implications and possible consequences of the findings are discussed in this paragraph. It relates partly to the introduction, where it is mentioned how the results of this study could be useful in the working environment.
The significant difference between the two regular conditions is of a greater magnitude than the difference between the favorable comparison and the control condition. This indicates that people attach more value to the issue whether the intrapersonal comparison is favorable or unfavorable, than to the issue whether an incentive is adjusted or not. However, the findings of the hypothesis test and the supplemental analysis are both significant and have both their implications.
4.2.1 Hypothesis test
The unfavorable intrapersonal comparison between the adjusted and unadjusted incentive is perceived as less fair than the favorable intrapersonal comparison between the incentives, while both situations contained the same RI-ratio. The t-test between the two regular experiment groups provided the result of the significant difference. This finding supports the prior literature and the study’s hypothesis. The presence of the self-interest bias regarding the intrapersonal comparison is confirmed.
This leads to the first conclusion, which is that organisations have to be cautious when they plan to make a downward subjective adjustment of an incentive. To prevent a downward adjustment, its probability has to be taken into account when an incentive plan is being produced. For example, an optimistic incentive plan can result in an increase of the employee’s motivation during a certain process, since the IFP is strongly associated with motivation (Bol & Smith, 2011; Bregn, 2013). However, when the incentive is adjusted downwards after the process, the IFP and therefore the motivation will decrease in the next process. This notion of caution is supported by Bol and Smith (2011), who mention that the use of discretion to correct an incentive downwards would likely to be considered as inappropriate and unfair, whether it is actually justified or not.
distributive justice is confirmed. So a downward adjustment has to be avoided at all times to prevent a decrease of the IFP due to an unfavorable intrapersonal comparison. An incentive plan related to a certain measure has to be reconsidered when it turns out that this plan could lead to a downward adjustment of the related incentive. The only option that remains when the incentive is too high and a downward adjustment has to be avoided, is paying too many unfounded incentives. This has neither positive consequences for an organisation.
4.2.2 Supplemental analysis
The supplemental analysis reveals that the mean IFP of the control condition, where no intrapersonal comparison was relevant, is significantly higher than the mean IFP’s of both regular conditions with the (un)favorable comparison. This finding suggests that a subjective adjustment is synonymous with a decrease of the IFP.
The lower IFP’s in the regular conditions contradict the findings of several studies, which concentrates on the positive effects of upward adjusted incentives on fairness perceptions. Subjectivity would enhance the accuracy of the performance evaluations and would therefore restore the fairness of the evaluations too (Bol 2008; Bol et al. 2015; Choon & Embi 2012).
However, there are also reasons to reconsider the positive effects of subjectivity. The experiment condition with the favorable comparison could be perceived as an over-reward, which, in turn, can result in a feeling of guiltiness (Williams, 1999). Besides, Lau and Buckland (2001) suggest that informal judgments lead to a decrease of the IFP. There is more reason to be doubtful about an incentive when the related evaluation is based on subjectivity. People could unconsciously have the feeling that the subjective adjustments result in inequity, because a deviation from the incentive agreement takes place. Inequity aversion, in turn, could lead to lower IFP’s (Fehr & Schmidt, 1999)
Furthermore, the controllability of the related performance measure could influence the IFP of the related outcome and incentive (Bol & Smith, 2011; Giraud et al. 2008). Giraud et al. (2008) find that people, who are held accountable for factors they do not control, have a decreased IFP regarding the related evaluation and incentive. In case of the control condition, the uncontrollability of the performance measure do not lead to a departure from the incentive agreement. This makes that it seems there is a high controllability of the performance measure in the control condition, which is strongly correlated with the IFP’s (Bol & Smith, 2011; Giraud et al. 2008). Therefore the participants may perceive the control condition’s case setting as fairer in general. It can be suggested that applying subjectivity to performance outcomes can restore the IFP of the related incentive to certain extent, but it will not reach the IFP of the unadjusted incentive, which is related to a controllable performance outcome.
This has consequences for organisations where subjectivity is applied regarding the subordinate’s incentive. For example, when supervisors consider to make a subjective adjustment of an uncontrollable performance outcome, they have to find out the subordinate’s perception of the unadjusted outcome at first. When the subordinate is not aware of any uncontrollability, there is no need to adjust the evaluation and the related incentive. Thus, supervisors have to be cautious regarding the adjustments of outcomes and incentives. Because when the performance outcome is once adjusted due to uncontrollability, the subordinate’s IFP will never reach the IFP of a subordinate who is not aware of any uncontrollability. Consequently, the controllability of a certain measure has to be sufficient when it is planned to be a performance measure.
In sum, this experiment shows that even an upward subjective adjustment of an incentive could be perceived as less fair than no subjective adjustment at all. This is probably the result of the quality of the related performance measure, which seems to be lower when an incentive is adjusted. This leads to a possible conclusion, which is that the IFP of an uncontrollable outcome cannot be completely restored by subjective adjustments. Because it is perceived as less fair than an controllable outcome that is not adjusted. Future research might provide the confirmation of this finding.
4.3 Limitations of the results
The validity of this research is strengthened by some factors in this research, such as the variety of the participants that represents the Dutch working population at the same time. Furthermore, the neutralization of the behavioural differences and the quite big sample size lead also to an improvement of the experiment’s validity. Despite that, the present research is not completely free of limitations.
Firstly, although the introduction of the questionnaire clearly explained what was expected from the participant, there was a limited control over to what extent the participants received the information as it was meant. For example, the empathy of the participant with the case setting has to be regarded as sufficient.
In section 3 of this research there is explained that the participants had no motives to bias their answers, because the case setting was only hypothetical and the individual scores would not be shared with other people. However, the fact that the case setting was only fictive, could also be a second limitation. It should be noted that people might be tempted to respond in a
Future research could extend the present study’s results by creating a real situation, which is comparable to the present study’s case setting.
Finally, there were no actual rewards for the respondents who participated in this experiment. This could have led to a lack of motivation to complete the questionnaire. On the other hand, the time and effort to fill in the questionnaire were minimal. With regard to the participants who did not fill in the questionnaire completely, I suppose that inattention would be the only cause here.
5 Summary and conclusion
In this study I investigated the people’s fairness perceptions related to the intrapersonal aspect of the distributive justice. Past research about people’s fairness perceptions revealed that the self-interest bias could play a determining role. However, these researches were concentrated on the interpersonal aspect of distributive justice. The few studies that focused on the intrapersonal variant, were combined with outcome satisfaction or other factors. On the contrary, the present study focuses on the intrapersonal variant in combination with the fairness perception of incentives.
The contribution of this paper is the combination of the concept of the intrapersonal distributive justice with the people’s fairness perceptions. Especially the focus on the favorability of the intrapersonal comparison regarding incentives is what makes this study unique. The intrapersonal comparisons are created by subjective adjustments of incentives. These subjective adjustments allow the participants of the experiment to make a comparison between the adjusted and unadjusted incentive.
I examined the influence of the favorability of the intrapersonal comparison on people’s fairness perceptions. It was argued that an upward adjusted incentive leads to a higher fairness perception than a downward adjusted incentive. This is despite the fact that the two subjectively adjusted incentives were actually equally fair, according to Adams’ equity theory. The difference in favorability of the intrapersonal comparisons leads to the expectation. This expectations corresponds with the individuals’ self-interest.
Through an experiment with only Dutch participants, I find support for the expectation and hypothesis. The results suggest that the intrapersonal comparison is used to determine the fairness perception in a self-serving way. People’s fairness perceptions differ from what is actually fair, according to the RI-ratio in Adams’ equity theory. Therefore the self-interest bias outweighs Adams’ equity theory. Referring to the hypothesis, the average fairness perception of an adjusted incentive is indeed higher when the intrapersonal comparison with the unadjusted incentive is favorable, than when the intrapersonal comparison is unfavorable for the individual. Individuals significantly consider a favorable intrapersonal comparison when they determine their fairness perception. This indicates that avoiding a downward subjective adjustment is the best option for an organisation to prevent a decrease of the subordinate’s fairness perception.
controllable performance measure, is more important for people’s fairness perceptions than a subjective adjustment of an incentive, which resulted from an uncontrollable performance measure.
This paper helps to develop understanding of how important an intrapersonal comparison can be for the influence of people’s fairness perception. A subjective adjustment of incentives can restore the people’s fairness perception, but this study explores also the negative consequences of subjectively adjusted incentives regarding the fairness perception.
The present study provides opportunities for future research. For example, future research could set up a real situation that is comparable to the present study’s case setting. Although there were taken several precautions to minimize the validity threats in this study, the rater bias was still present. When the case setting of this study becomes reality, the participants receive actual rewards, which may have as a result that there is less chance that people will react in politically correct manner.
Finally, the supplemental analysis of this study deserves more attention of future research. Researches so far concentrated on the controllability of performance measures in relation with the fairness perceptions. More controllability of the performance measure results in a higher fairness perception. Besides, it is studied that subjective adjustments could restore the fairness perceptions of an incentive. The supplemental analysis of the present study leads to a suggestion which combines the two aforementioned research findings: An unadjusted controllable performance measure results in a higher fairness perception, than a uncontrollable performance measure that is subjectively adjusted. This suggestion lays a foundation for future research to confirm.
6 References
Adams, J. S. (1965). Inequity in social exchange. Advances in experimental social psychology, 2(267 299).
Bellavance, F., Landry, S., & Schiehll, E. (2013). Procedural justice in managerial performance evaluation: Effects of subjectivity, relationship quality, and voice opportunity. The British Accounting Review, 45(3), 149-166.
Bol, J. C. (2008). Subjectivity in compensation contracting. AAA Management Accounting Section (MAS) 2006 Meeting Paper.
Bol, J. C., Hecht, G., & Smith, S. D. (2015). Managers' discretionary adjustments: The influence of uncontrollable events and compensation interdependence. Contemporary Accounting Research, 32(1), 139-159.
Bol, J. C., & Smith, S. D. (2011). Spillover effects in subjective performance evaluation: Bias and the asymmetric influence of controllability. The Accounting Review, 86(4), 1213-1230. Bos, van den, K., & Lind, E. A., Wilke, H. A. (2001). The psychology of
procedural and distributive justice viewed from the perspective of fairness heuristic theory. Utrecht University
Bregn, K. (2013). Detrimental effects of performance-related pay in the public sector? On the need for a broader theoretical perspective. Public Organization Review, 13(1), 21-35.
Cappelen, A. W., Hole, A. D., Sørensen, E. Ø., & Tungodden, B. (2005). The pluralism of fairness ideals: An experimental approach.
CBS (2015). Beroepsbevolking. Geraadpleegd op 21 april 2016, van https://www.cbs.nl/nl -nl/achtergrond/2015/49/beroepsbevolking.
CBS Statline (2016). Werkzame beroepsbevolking; vergrijzing per bedrijfstak SBI 2008. Geraadpleegd op 21 april 2016, van http://statline.cbs.nl/StatWeb/publication/?VW=T&DM=SLN L&PA=80832NED&LA=NL
Choon, L. K., & Embi, M. A. (2012). Subjectivity, organizational justice and performance appraisal: understanding the concept of subjectivity in leading towards employees’ perception of fairness in the performance appraisal. Procedia-Social and Behavioral
Cole, N. D., & Flint, D. H. (2004). Perceptions of distributive and procedural justice in employee benefits: flexible versus traditional benefit plans. Journal of Managerial
Psychology, 19(1), 19-40.
Downing, S. M., & Haladyna, T. M. (2004). Validity threats: overcoming interference with proposed interpretations of assessment data. Medical Education, 38(3), 327-333.
Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation. Quarterly journal of Economics, 817-868.Field, A. (2009). Discovering Statistics using SPSS. London: Sage Folger, R. (1986). A referent cognitions theory of relative deprivation. In Relative deprivation and
social comparison: The Ontario symposium (Vol. 4, pp. 33-55).
Giraud, F., Langevin, P., & Mendoza, C. (2008). Justice as a rationale for the controllability principle: A study of managers’ opinions. Management Accounting Research, 19(1), 32-44. Greenberg, J. (1986). Determinants of perceived fairness of performance evaluations. Journal of
applied psychology, 71(2), 340.
Hartmann, F., & Slapničar, S. (2012). Pay fairness and intrinsic motivation: the role of pay transparency. The International Journal of Human Resource Management, 23(20), 4283-4300. Huffman, C., & Cain, L. B. (2001). Adjustments in performance measures: Distributive and
procedural justice effects on outcome satisfaction. Psychology & Marketing, 18(6), 593-615. Ittner, C. D., Larcker, D. F., & Meyer, M. W. (2003). Subjectivity and the weighting of
performance measures: Evidence from a balanced scorecard. The Accounting Review, 78(3), 725-758.
Jawahar, I. M., & Stone, T. H. (2011). Fairness perceptions and satisfaction with components of pay satisfaction. Journal of Managerial Psychology, 26(4), 297-312.
Lau, C. M., & Buckland, C. (2001). Budgeting—The role of trust and participation: A research note. Abacus, 37(3), 369-388.
Loseman, A., Miedema, J., Van den Bos, K., & Vermunt, R. (2009). Exploring how people respond to conflicts between self-interest and fairness: Influence of threats to the self on affective reactions to advantageous inequity. Australian Journal of Psychology, 61(1), 13-21. Maas, V. S. (2011). Writing an MSc thesis in Management Accounting. Erasmus School of Economics.
