1 The search for mediators in and moderators on the relationship between Leader-
Member Exchange (LMX) and in-role job performance of an employee
Master Thesis
Msc Human Resource Management Faculty of Economics and Business
University of Groningen
June, 2017
Willem-Jan la Roi
Student number: 2397145 Parkzijde 9, 9713WE Groningen
Tel.: +31610630432 E-mail: w.j.la.roi@student.rug.nl
Supervisor Dr. Ramzi Said
2 ABSTRACT
In order to enhance employee’s in-role job performance, there is an important role of the leader. Leader-Member Exchange (LMX) is a style of leading with, as is widely supported, potential to influence in-role job performance of an employee positively. As many researchers mention, the phenomenon of reciprocity is strongly related to LMX theory. Strangely enough, however, up till now hardly any research has been devoted to study whether reciprocity is indeed the explanation why employees perform better as a result of an LMX relationship between leader and employee. This study has focused on filling this research gap. In addition, we have examined ‘workload’ as moderating variable on the relation between positive
reciprocity and an employee’s in-role job performance. Results of our sample of 366
employees and 53 supervisors show no support for our main expectations. Our most important findings are that positive reciprocity does not operate as mediator, and that workload does operate as moderator, but (surprisingly) positively. A discussion of these unexpected findings is given with several possible explanations and direction for future research.
3 TABLE OF CONTENTS
ABSTRACT ... 2
INTRODUCTION ... 4
THEORY ... 8
LMX and employee performance ... 8
The mediating effect of positive reciprocity ... 11
The moderating effect of workload ... 16
METHOD ... 19
Procedure and sample ... 19
Measures ... 20
Analytic approach ... 21
RESULTS ... 22
Correlations ... 22
Hierarchical Regressions... 22
Additional analysis ... 24
DISCUSSION ... 24
Summary of results... 24
Theoretical and managerial implications ... 25
Limitations and future directions ... 30
CONCLUSION ... 32
References ... 33
Appendix ... 40
TABLE 1: DESCRIPTIVES EMPLOYEES ... 40
TABLE 2: DESCRIPTIVES SUPERVISORS ... 41
TABLE 3: CORRELATION TABLE ... 42
TABLE 4: HIERARCHICAL REGRESSION TABLE (Mediation) ... 43
TABLE 5: HIERARCHICAL REGRESSION TABLE (Moderation) ... 44
FIGURE 2: INTERACTION-PLOT ... 45
4 INTRODUCTION
In 2016, around 1,6 million companies are registered in the Netherlands (Centraal Bureau voor Statistiek, 2016). In a country of 17 million people, it can be said that a lot of competition occurs between firms. So, organizational survival is a challenge in itself. Foster
& Kaplan (2001) for example showed that 61 percent of companies that were in the Forbes Top 100 in 1917, no more existed in 1987. An important factor related to organizational survival, is employee’s in-role job performance. In-role job performance can be defined as all activities related to the formal job requirements of an employee (Borman & Motowidlo, 1997). The importance of in-role job performance of employees for an organization can be explained by the consequences of underperformance. Underperformance has several negative consequences for a company. First, and of predominant influence, underperformance costs an organization money. For example, performing or underperforming of football players for a football team participating in a highly lucrative league may be the difference between winning high prices or winning nothing. Also, underperformance of an employee can lead more
quickly to turnover (de Dios Tena & Forrest, 2007). Turnover mostly leads to hiring new people which costs time and money (Karsan, 2007). Then, a high turnover rate can take a big chip out of the finances of a company. So, failure may obstruct financial advantages in several ways (costs, consequences of good marketing position, et cetera) (de Dios Tena & Forrest, 2007). In stimulating the in-role job performance of an employee (which we will name
‘employee performance’ further in this study), there is an important role of a leader (Yukl, 2013). In previous research, several leadership constructs have been related to employee performance (Kuoppala et al., 2008; Yukl, 2013). Leader-Member Exchange (LMX) is one of these. There is a lot of support that LMX is related to employee performance (for example:
Graen & Uhl-Bien, 1995). However, there are still unanswered questions about this theory.
Therefore, we focus this study on LMX.
5 The LMX theory can be distinguished from other leadership theories in its
accentuation on the dyadic level; the relation between leader and employee (Dansereau, Cashman, & Graen, 1973; Dansereau, Graen, & Haga, 1975; Gerstner & Day, 1997; Graen &
Cashman, 1975). LMX can be defined as the physical and mental effort, material resources, information and emotional support which is exchanged between a leader and an employee (Liden, Sparrowe & Wayne, 1997). LMX can be seen as a construct which mainly has been measured in terms of ‘quality of LMX relationship’. Previous research has shown that the quality of an LMX relationship between leader and employee is of large importance regarding the productivity and performance of an employee (Gerstner & Day, 1997). There is meta- analytic evidence that this LMX relationship quality is related to job performance (Gerstner &
Day, 1997). High quality LMX relationships are marked by mutual trust, respect and commitment. Low quality LMX relationships are characterized by formal, role-defined interactions resulting in one-way influence from leader to employee (Gerstner & Day, 1997;
Graen & Uhl-Bien, 1995; Janssen & Van Yperen, 2004). Although LMX has been
conceptualized and refined many times (Graen & Uhl-Bien, 1995; Walumbwa, Cropanzano &
Goldman, 2011), research around LMX can gain further theoretical refinements. It is widely accepted that LMX leads to higher employee performance, but we do not know precisely why, and under what circumstances this relationship exists (Walumbwa, Cropanzano &
Goldman, 2011). Addressing this research gap, this study examines the impact of a mediator and a moderator.
So, LMX is about relations. This relational approach is based on the social exchange theory. The social exchange theory as stated by Blau (1964) refers to social exchanges in a relationship as a result of a certain unspecified cost-benefit analysis together with a
comparison of alternatives; doing another a favour has automatically a certain expectation in it referring to some future return. This principle seems to be really close in basics to the
6 phenomenon of positive reciprocity. Positive reciprocity can be defined as the impulse or intrinsic desire to be kind to those who have been kind to us (Fehr & Gächter, 1998; McCabe, Rigdon & Smith, 2003). So, the idea of reciprocity is also based on an exchanging relation.
This exchanging relation includes a willingness of an employee to do something extra for his or her leader (Liden & Graen, 1980). For this reason, we expect that the relationship between LMX and employee performance can be explained by the phenomenon of positive reciprocity.
Hereby, we not only test just a mediating effect, we examine a fundamental assumption of the LMX theory, which not has been examined yet in literature.
In addition, this study looks to the moderating effect of workload on the relationship between positive reciprocity and employee performance. In this study, workload is defined as the amount of work an employee has to get done. We look to the moderating effect of
workload, because workload is a variable which is relevant in every organization, and for we expect workload to have a significant influence on the relation between positive reciprocity and employee performance. Namely, a high workload can lead to both psychological and physical demands for employees over sustained periods of time (Schaufeli, Bakker & van Rhenen, 2009). As a result, a high workload can lead to mental depletion, a higher level of stress and a certain disruption of performance (Boksem & Tops, 2008; Hockey, 1997).
Therefore, we expect that a high workload weakens the relationship between positive
reciprocity and employee performance. Together, our expected relations form the conceptual model as shown in figure 1.
7 FIGURE 1: Expected relations
As already has been stated, more knowledge about context variables is necessary around the relation between LMX and performance (Walumbwa, Cropanzano & Goldman, 2011). In this study, we address this research gap by examining positive reciprocity as a mediator and workload as a moderator. Several studies have explored the concept of LMX (for example: Dienesch & Liden, 1986; Liden, Sparrowe & Wayne, 1997) and positive reciprocity (for example: Fehr & Gächter, 1998; McCabe, Rigdon & Smith, 2003). Both constructs have an important foundation in the social exchange theory. However, positive reciprocity has not been tested yet whether it operates as a mediator in the relation between LMX and employee performance. Therefore, this study contributes a fundamental
examination of LMX theory to existing research around LMX theory. Besides, by examining the moderator workload on the relation between positive reciprocity and employee
performance, we also contribute to existing research around LMX theory, for there is a gap in knowledge about contextual variables under which LMX, positive reciprocity and employee performance are related. Besides that this study addresses an important research gap, it is also practical relevant. By examining the mediating role of positive reciprocity and the moderating role of workload, quite relevant recommendations can be made to leaders. Namely, by having
8 insight about contextual factors influencing the performance of employees, a leader can adjust his or her style of leading in favour of increasing his or her employee’s performance.
Concretely, we can advise leaders about when to adjust the level of workload in order to make the employees perform better.
This thesis is divided into four sections. First, the theoretical foundations arising from the literature about leader member exchange, positive reciprocity, workload, employee
performance and the stated relations among these variables are examined. Second, the method section is stated. In this section, all information about the large field study we conducted in the Netherlands among employees and their direct supervisor. Third, results from the data- analysis are given and hypotheses are tested. In the final section, our conclusions are discussed, practical implications are explored and directions for future research are given.
THEORY LMX and employee performance
In the early ‘70’s, a new leadership theory was introduced; the LMX theory. As named in the introduction, the LMX theory can be distinguished from other leadership theories in its accentuation on the dyadic level (Dansereau, Cashman, & Graen, 1973; Dansereau, Graen, &
Haga, 1975; Gerstner & Day, 1997; Graen & Cashman, 1975). This dyadic level means a two- way exchanging relationship between leader and employee. The theory of LMX has its foundation in two theories; the social exchange theory and the role theory (Liden & Maslyn, 1998). The principle of the social exchange theory explains why doing another a favour has automatically a certain expectation in it referring to some future return (Wayne, Shore &
Liden, 1997). The role theory suggests that all everyday behaviours and activities can be analysed as an outcome of a certain role (for example: father, leader, teacher) (Graen, 1976).
To clarify LMX theory in more detail, these two theories are elaborated in the next
9 paragraphs. In the subsequent paragraphs our main reasons for expecting the relation between LMX and employee performance are elaborated.
The social exchange theory as stated by Blau (1964) refers to social exchanges in a relationship as a result of a certain unspecified cost-benefit analysis together with a
comparison of alternatives. Doing another a favour has automatically a certain expectation in it referring to some future return. In this, the social exchange theory provides an important fundament for LMX (Sparrowe & Liden, 1997). The characteristic dyadic level of
relationship of LMX is in itself a social exchange, wherein both parties have to provide something of value to the other party in an equal and fair context (Green & Scandura, 1987).
An important factor of social exchange is mutual liking (Wayne et al., 1997). Liking
influences the development of the relationship between leader and employee and is a predictor of LMX (Dockery & Steiner, 1990; Liden, Wayne & Stilwell, 1993; Wayne & Ferris, 1990).
Therefore, an employee liking a leader is related to a more positive view of the quality of the LMX relationship from the perspective of the employee (Wayne et al., 1997).
The role theory can be explained on the hand of the development of an LMX
relationship. The theory of LMX suggests that leaders do not use the same style in leading all employees, but develop a different, specific, type of relationship with each employee
(Dansereau, Graen, & Haga, 1975; Graen & Cashman, 1975; Graen, Novak, & Sommerkamp, 1982; Graen & Scandura, 1987; Liden & Graen, 1980). However, according to Bauer and Green (1996), all types of relationships go through a development with three role defining stages. The first stage is named ‘role taking’. In this stage, the relationship between leader and employee is highly influenced by demographic characteristics and personalities of both leader and employee (Dienesch & Liden, 1986; Mayer et al,. 1995). Bauer and Green (1996) suggest this is because of the limited information a leader and employee has about the other early in the relationship. In time, the relationship develops and further exchanges are made. Then, the
10 relationship is in the second stage according to Bauer and Green (1996): the role making stage. This is the stage in which the relationship is characterised by specific behaviours, resulting from a certain obtained level of trust. For example, the specific level of trust of a leader to an employee makes a leader taking risk by delegating work to a employee. Lewis &
Weigert (1985) describe this stage as the stage wherein cognitions are put into action. After these two stages, the relationship between leader and employee has developed to a level where behaviours of a leader and employee become more predictable (Graen & Scandura, 1987). This stage is called role routinization. The role routinization is characterised by the relationship becoming emotional, based on affect (Graen & Scandura, 1987). A certain emotional bond has been established in which the relationship can now be measured in terms of quality of LMX relationship. Then, as named in the introduction, there is a range between developed high-quality exchange relationships and developed low-quality exchange
relationships.
High quality exchange relationships are marked by mutual trust, respect and
commitment, leading to two-way influence between leader and employee (Graen & Uhl-Bien, 1995; Janssen & Van Yperen, 2004). This kind of relationship is often accompanied by a leader providing to the employee special privileges (like access to key information), career- enhancing opportunities (like special work assignments) and a higher discretion in doing their jobs (Graen & Uhl-Bien, 1995; Wayne et al., 1997). As a result, employees are more
motivated to do something extra for their leaders. For example, these employees may, besides their own tasks, voluntarily work overtime to pass certain deadlines (Graen & Uhl-Bien, 1995). This is supported by Liden and Graen (1980): employees having a high-quality relationship with their leaders make more extra contributions beyond their formal job. So, a high-quality relationship consists of a form of preferably treatment of the leader together with a higher motivation to perform of the employee (Liden & Maslyn, 1998).
11 In contrast, employees having a low quality LMX relationship only perform their own tasks (Liden & Graen, 1980). Low quality exchange relationships are marked by formal, role- defined interactions resulting in one-way influence from leader to employee (Graen & Uhl- Bien, 1995; Janssen & Van Yperen, 2004). These kind of relationships are therefore seen as
‘contractual’; give a little, take a little. This is often accompanied by a leader exercising formal authority and allocating standard advantages in return for standard job performance (Graen & Uhl-Bien, 1995). Where in a high-quality relationship an employee is seen as
‘trusted assistant’, in a low-quality relationship an employee is seen as ‘hired hand’. In other words, a differentiation between high quality LMX relationships and low quality LMX relationships is visible at both the level of allocated resources as well as the level of trust between leader and employee. Therefore, it can be said that employee performance is a form of currency in the quality of the social exchange between leader and employee.
Previous research about LMX theory already suggested that the level of quality of the exchanging relationship that arise between leader and employee are of predictive value for the performance of an employee (Gerstner & Day, 1997; Graen & Uhl-Bien, 1995; Janssen &
Van Yperen, 2004). We also expect to find this relation in this study, because of the two main reasons we elaborated. First, the motivation of an employee to do something extra for the leader the employee trusts, respects and commits to. Second, the fact that employees having a high LMX relationship receive more task-related resources to do their work, and thus perform better. Therefore, we hypothesize:
Hypothesis 1: LMX is positively related to employee performance.
The mediating effect of positive reciprocity
As many researchers mention (for example: Dienesch & Liden, 1986; Graen & Uhl-Bien, 1995), social exchange theory and reciprocity are foundations of LMX theory. Strangely enough, however, up till now hardly any research has been devoted to study whether
12 reciprocity is indeed the thing why employees perform better as a result of LMX. Therefore, in the present paper, we test this fundamental assumption. We elaborate this mediating effect of positive reciprocity by first showing support for the relation between LMX and positive reciprocity, then the relation between positive reciprocity and employee performance.
The phenomenon of reciprocity
Previous research has shown that a large majority of people have a certain willingness to reward received gifts. This willingness can be seen as an intrinsic impulse. This is actually the case in all fields of life (Fehr and Gächter, 1998). Studies about both humans and animals show that positive reciprocity is a quite common part of behaving together with others. For example, De Waal (1991) showed in a study about food sharing among chimpanzees that chimpanzees share their food with other chimpanzees depending on the past interactions.
From many other studies becomes clear that even when people do not know each other and do not see with whom they interact, people tend to reciprocate positively. This is shown by for example Fehr et al. (1993). In their gift exchange game player 1 could give voluntarily resources to player 2, which could be seen as a gift because player 2 didn’t have to give something in return for it. After, player 2 could also transfer certain resources to player 1. The big majority of players 1 made transfers of resources to player 2 which in nearly all cases were rewarded by a gift from player 2. What becomes clear of this, is that most people have an intrinsic impulse to do something back to others who have done something to them.
As stated before, LMX theory is based on a dyad linkage which consists of two-way exchanging behaviour (Wang et al., 2005; Uhl-Bien, 1995). So in essence, LMX is about social exchanging. Social exchanges are based on a trust that behaviours of goodwill will once be reciprocated (Settoon, Bennett & Liden, 1996). According to Graen and Uhl-Bien (1995), then, time makes exchanges between leader and employee result in reciprocation. This can be explained on the hand of the role theory translated to the development stages of an LMX
13 relationship as mentioned before. In stages one and two, trust and respect are important
factors which partly determine the quality of LMX relationship in stage three (Bauer & Green, 1996). These two factors (trust and respect) have to be built up over time. This is supported by Walumbwa, Cropanzano & Goldman (2011); high quality LMX relationships are long-term relationships. So, LMX results in a development of social exchanges between leader and employee, which are followed by positive reciprocity due to an intrinsic impulse.
Besides this development of two-way exchanging, another theoretical explanation for the expected positive link between LMX and positive reciprocity is the factor ‘personal identification’. The more a leader exchanges with an employee, the more an employee will internalize the vision and behaviours of a leader: the employee will feel personal identified (Kelman, 1958; Wang et al., 2005). This process may also lead to a more collective thinking of an employee instead of thinking of self-interest which in its turn leads to more
encouragement and recognition of the leader (Wang et al., 2005). In return, an employee answers the encouragement of the leader by showing more commitment. In this, he is reciprocating the leaders’ behaviours (Wang et al., 2005; Yukl, 2012). So, LMX results in personal identification of a leaders’ vision and behaviours of an employee, which causes positive reciprocal behaviours of an employee.
Elaborating the differences between high- and low-quality LMX relationships referring to positive reciprocity, we come to the same expected positive relation between LMX and positive reciprocity. As mentioned, it is expected that a high-quality relationship leads to positive reciprocity of an employee. On the other hand, low-quality exchange relationships are marked by formal, role-defined interactions resulting in one-way influence from leader to employee (Graen & Uhl-Bien, 1995; Janssen & Van Yperen, 2004). These relationships consist of less trust (Graen & Uhl-Bien, 1995), while trust is seen as very important antecedent for reciprocical behaviours (Settoon, Bennett & Liden, 1996).
14 Therefore, in these relationships consisting of less trust, reciprocal behaviours are less likely to be shown.
Summarized, the basics of LMX theory together with the social exchange theory, the intrinsic impulse of reciprocity and the factor ‘personal identification’ support our second hypothesis:
Hypothesis 2: LMX is positively related to positive reciprocity of an employee.
Positive reciprocity as a driver for performance
As we elaborated in the previous section, a chimpanzee is in most cases willing to reward a banana if he or she has received one. This phenomenon of reciprocity is also in the work setting a thing (Fehr & Gächter, 2000; De Dreu & Nauta, 2009). In this paragraph, we show theoretical support for our expectation that positive reciprocity also leads to performance of an employee.
Our main argument is about the level of motivation of an employee. Positive
reciprocity is positively related to increased motivation (Graen & Uhl-Bien, 1995). Therefore, positive reciprocity can operate as a powerful motivator. Previous research has showed that almost all people can be driven by positive reciprocity (Fehr & Gächter, 1998). Fehr &
Gächter (1998) support this driving effect of positive reciprocity: employees showing a lot of positive reciprocal behaviours are more motivated to give extra effort on top of what is expected by financial performance incentives. On the other hand, employees scoring low on positive reciprocity are only driven by financial performance incentives. Besides, Fehr and Gächter (1998) state: explicit financial performance incentives reduce the willingness to reward voluntarily. An explanation for this is that explicit financial performance incentives may create a work atmosphere of distrust, where positive reciprocating behaviour is based on trust (Fehr & Gächtner, 1998). So, the shown increased motivation of an employee scoring
15 high on positive reciprocity makes the difference between giving effort just as what is
expected, or giving extra effort, more than as expected.
Secondly, studies of Fehr & Gächter (1998; 2000) show that employees scoring high on positive reciprocity have another focus than employees scoring low on positive reciprocity.
This difference in focus is about employees focused on self-interest and employees focused on other-interest. This differentiation makes sense, for existing research has shown these different focuses to have a different outcome regarding employee performance. In elaborating the phenomenon of reciprocity, Fehr & Gächter (2000) have made a link between the self- interest versus other-interest and high positive reciprocity versus low positive reciprocity.
People scoring high on positive reciprocity are linked to having a focus on other-interest, where people scoring low on positive reciprocity are linked to having a focus on self-interest.
About these constructs, De Dreu & Nauta (2009) have found that employees scoring high on other-interest score better on job performance than employees who score high on self-interest.
They also suggest that this can be explained by the level of motivation of employees and the fact that employees scoring high on positive reciprocity get more privileges and discretion.
Namely, as a result, these employees are willing to do something extra for their leader (De Dreu & Nauta, 2009). This difference in outcomes between employees scoring high and employees scoring low on positive reciprocity strengthens our expectation that positive reciprocity is positively related to employee performance.
Furthermore, as we elaborated, positive reciprocity is a kind of intrinsic impulse in human being to do something for the person who has done something to us. This intrinsic aspect makes it likely that positive reciprocity is automatically positively related to employee performance. Namely, the vision and good behaviours of a leader are kind of automatically reciprocated due to this intrinsic aspect. Therefore, we expect that positive reciprocity leads to performance of an employee.
16 In short, the principle of reciprocity can be seen as part of human being. In work setting, employees scoring high on positive reciprocity are more motivated to perform than employees scoring low on positive reciprocity. Therefore, hypothesis three is as follows:
Hypothesis 3: Positive reciprocity of an employee is positively related to employee performance.
We have showed that there is support to expect the relations of LMX and employee performance, LMX and positive reciprocity as well as positive reciprocity and employee performance. Together these expectations form a relevant framework. Combining the above reasoning, we expect that positive reciprocity has a mediating effect on the relationship between LMX and employee performance. Therefore, hypothesis 4 is:
Hypothesis 4: The positive relationship between LMX and employee performance is mediated by positive reciprocity of an employee.
The moderating effect of workload
As named in the introduction, there are unanswered questions about LMX theory, as what moderators play a role in the relationship between LMX and employee performance (Walumbwa, Cropanzano & Goldman, 2011). To extend our mediation test of positive reciprocity, and for we have expectations for workload to operate as a moderating variable, it is necessary examine this moderating effect. The expected moderating effect of workload arises from some underlying factors. These are elaborated in subsequent paragraphs.
Our main reason for our expectation of workload to operate as moderator, is that a high workload has been associated with mental depletion (Van der Hulst & Geurts, 2001).
Over the last decades, the content of work has changed largely from physical to mental. This has entailed a significant increase in complaints related to mental depletion (Boksem & Tops, 2008). Research in the Netherlands shows that in the period 1993-2008, the percentages men
17 and women complaining about their mental depletion have increased from 24% to 33% (men) and from 38% to 50% (women) (Boksem & Tops, 2008). Several studies report comparable results in other countries (see for example Ricci et al., 2007). Theoretically a high workload can lead to both psychological and physical demands for employees over sustained periods of time (Schaufeli, Bakker & van Rhenen, 2009). Mental depletion has been related with
reduced cognitive and behavioural performance (Boksem et al., 2005; Lorist et al., 2005).
Especially the cognitive impairments are in this case relevant. Namely, Van der Linden et al.
(2006) show that depleted employees have difficulties in focusing and planning. In addition, Boksem et al. (2006) found that employees complaining about being mentally depletion have difficulties in preparing their responses and ignoring irrelevant information. All these
impairments (focusing, planning, preparing responses and ignoring irrelevant information) make it likely that an employee who scores high on showing positive reciprocal behaviours, does not perform at its best. So, we expect that a high workload weakens the relation between positive reciprocity and employee performance, because of the workload-mental depletion relationship. We can strengthen this thought by means of two examples, wherein stress and mental depletion are weakening the relation between positive reciprocity and employee performance.
The first example is about the intention to quit of an employee. At first, the intention to quit with a job is influenced by feelings of stress (Firth, et al., 2004). It has been shown that stress is being influenced by job stressors. One important job stressor is a high workload (Firth et al., 2004). So, a high workload may be positively related to a faster intention to quit with the job. Firth et al. (2004) show that on its turn, an intention to quit is seen as a strong indicator for ‘quitting behaviour’: issues referring to a lower employee performance.
Employees showing such quitting behaviour are more busy with thinking about quitting and other negative thoughts about their work. It might be that the time and energy these thoughts
18 take, inhibit the performance resulting from positive reciprocity. So despite wanting to do something back for the leader, the employee has his mind somewhere else.
Another example is about absenteeism. A high workload is positively related to stress which is positively related to absenteeism (Hockey, 1997). As a result, an employee would simply have lack of time to reciprocate all the leader’s good behaviours. So the willingness to reciprocate positively will not to its fullness lead to a higher performance because of the time and energy which are away due to a high workload. As a result, the positive relationship between positive reciprocity and employee performance would be weaker. Considering these examples in which mental depletion and stress, as functions of workload, might weaken the relationship between positive reciprocity and employee performance, we state our hypothesis 5 is as follows:
Hypothesis 5: Workload moderates the relationship between positive reciprocity of an employee and employee performance, such that this relationship is weaker under high
workload, and stronger under low workload.
Taking all these hypotheses together, we come to a sixth hypothesis, which actually is the aggregation of all expected relations (see for the overview of these relations together the proposed model in Figure 1, introduction). We test a mediating effect of positivity and a moderating effect of workload, therefore we come to a moderated-mediation model. So, hypothesis 6 is as follows:
Hypothesis 6: The mediation effect of positive reciprocity of an employee in the relation between LMX and employee performance is weaker under high workload, and stronger under low workload.
19 METHOD
In this section, we elaborate how we have performed this research. At first, we describe the whole procedure of this study and the composition of our sample. Then, we explain what measurements have been used measuring the variables of this study. Lastly, we name what statistical analyses we have done to analyse the data.
Procedure and sample
Data have been collected by surveys. Respondents were employees with their direct supervisor. These respondents have predominantly been acquired by networking; we
approached organizations by using the personal network. The range of types of organizations was large, as organizations taking part in this study are from governmental institutions, the pharmaceutical industry, the construction industry, et cetera. Of the companies which wanted to participate in our study, we collected all of the email addresses of the respondents. Then, the supervisors received an information letter about the study. In this letter, the surveys were announced and the confidentiality of their response was assured. After this, the respondents received the survey. There were two different surveys, separately for employees and their direct supervisors. Supervisors were asked to rate their employees on employee performance, whereas subordinates were asked to rate the quality of LMX relationship, their level of positive reciprocity and their workload.
In total, of the 480 approached employees, 366 completely filled in their survey, which makes an overall response rate of 76 %. Of this response rate percentage, there were 154 male employees and 212 female employees, which makes a ratio men/women of 42/58. The
average age was 43 years (SD = 11,8) and the average years of work experience was 21 (SD = 11,6). The distribution of education level was 1 employee with a PhD, 45 employees with a university degree (13 %), 143 employees with an HBO degree (39 %), 147 employees with an MBO degree (40 %), 29 employees with a primary school degree of education (8 %) and 1
20 employee with no degree of education. For an overview of these frequencies, see table 1 (Appendix).
Of the 61 approached supervisors, 53 completely filled in the survey, resulting in a response rate of 90 %. The sample of supervisors consisted of 30 men (57 %) and 23 women (43 %). The mean age was 46 years (SD = 10,3), and the average years of work experience was 24 (SD = 10,4). Of the supervisors, 15 had a university degree (28 %), 30 had an HBO degree (57 %), 6 had an MBO degree (11 %) and 2 had a primary school degree (4 %). For an overview of these frequencies, see table 2 (Appendix).
Measures
The four constructs of our model (LMX, positive reciprocity, workload and employee performance) have been measured with a 5-point Likert scale with response options ranging from 1 ‘strongly disagree’ to 5 ‘strongly agree’. All used measures are from existing
literature.
Leader member exchange has been measured by means of the scale of Liden & Maslyn
(1998). The Cronbach’s alpha of this construct was 0.87. The scale measured four dimensions of leader member exchange, with the following reliabilities: contribution (two items, a = .58), affect (three items, a = .86), loyalty (three items, a = .84), and professional respect (three items, a = .88). One example item of this instrument, measuring the dimension ‘contribution’
is ‘I do work for my supervisor that goes beyond what is specified in my job description’.
Positive reciprocity. We used the scale of Perugini, Gallucci, Presaghi, & Ercolani (2003) to
measure positive reciprocity. Sample items include ‘If my supervisor does me a favor, I am prepared to return it’ and ‘I go out of my way to help my supervisor who has been kind to me before’. The three items measured for this construct showed together a Cronbach’s alpha of 0.74
21 Workload. We used the scale of Van Yperen & Hagedoorn (2003) to measure the level of
workload of an employee. This instrument consists of five items, with a reliability (the stated Cronbach’s alpha) of 0.90. One sample item is: ‘I work under time pressure’
Employee performance has been tested with the scale of Wayne, Shore & Liden (1997). The
instrument included three items. Cronbach’s alpha was for this construct 0.86. One sample item of this instrument is ‘...performs his/her duties as I like to see them performed’.
Control variables were included to increase accuracy to our results. We have used ‘gender’
and ‘age’ as control variables. Gender has been measured as a dichotomous variable (1 was male, 2 was female). Age has been measured by asking an open question.
Analytic approach
In order to explore all proposed relationships separately, we have calculated the correlations.
Then, to examine the mediating effect, we have done a hierarchical regression analysis.
For mediation, we have established four steps. These four steps are equal to our four hypotheses; subsequently we examined the relation between LMX and employee performance (1), LMX and positive reciprocity (2) positive reciprocity and employee performance (3) and LMX, positive reciprocity and employee performance together (4). To state whether there was mediation, we also conducted the Sobel test, which is an analytical approach of Sobel (1982).
This is a test which provides a significance test for the indirect effect of the independent variable (IV) on the dependent variable (DV) via the mediator.
For moderation, we have established three steps. In this, we have calculated the interaction effect using standardized variables (according to Aiken & West, 1991). First, we examined the relation between positive reciprocity and employee performance (1). Second, we calculated the relation between positive reciprocity, workload and employee performance (2). Third, we examined the regression between positive reciprocity, workload, the interaction between positive reciprocity and workload, and employee performance. For testing the
22 complete model, we made use of model 14 of the SPSS tool of PROCESS designed by Hayes (2013).
RESULTS
In this section, we provide the outcomes of the statistical tests we have conducted as described in the methodology section. The tables and figure we refer to can be found in the Appendix.
Correlations
Table 3 shows the correlations between all variables (LMX, positive reciprocity, workload, employee performance, age and gender). From this matrix, it appears that LMX is positively correlated to employee performance (r = .14, p < .01). Furthermore, LMX is positively correlated to positive reciprocity (r = .27, p < .01). LMX is also correlated to our moderator workload (r = -.11, p < .05), although this correlation is negative. Positive reciprocity is not significantly correlating with both workload (r = -.01, p > .10) as well as employee
performance (r = -.03, p > .10). Lastly, there is a marginally significant correlation between workload and employee performance. This correlation is negative (r =- .10, p < .10).
Considering the control variables, gender does not correlate with all of the other variables. Age, however, appears to negatively correlate on positive reciprocity (r = -.20, p <
.01), employee performance (r = -.20, p < .01) and positively correlate on workload (r = .09, p < .10). So, the older an employee is, the less positive reciprocal behaviours he or she shows, the less he or she performs and the more workload is experienced.
Hierarchical Regressions
The second way of examining the mediation and moderation effect, was through hierarchical regression. Table 4 shows the results of the regressions in the mediation part of the model,
23 table 5 shows the results of the regressions regarding to the test of the moderating effect of workload. In both tables the B-values are given, together with their standard errors.
Mediation. From table 4, it appears that there is a significant positive relationship between
LMX and employee performance (B = .13, p < .05). Therefore, hypothesis 1 can be accepted.
LMX appears also to be significantly related to positive reciprocity (B = .35, p <.01). In addition, hypothesis 2 can be accepted. Then, our results indicate no significance between positive reciprocity and employee performance (B = -.06, p > .10). This makes that we reject hypothesis 3. Because this insignificance on the relation between positive reciprocity and employee performance, there can not be a mediating effect of positive reciprocity. Therefore, hypothesis 4 has to be rejected automatically. However, the relationship between positive reciprocity and employee performance becomes significant when LMX is included. This becomes clear from both table 4 (B = -.09, p < .05) as well as the Sobel test we did (statistics:
Ζ = -2.05, p < .05, two-tailed). This indicates that the relation between LMX and positive reciprocity is quite strong, because this relation makes the relation between positive reciprocity and employee performance significant. Considering the control variables, it appears that gender does not predict employee performance nor positive reciprocity.
However, in line with the correlations, age seems to be a strong predictor for both positive reciprocity (B = -.01, p < .01) as well as employee performance (B = -.01, p < .01).
Moderation. Table 5 shows the results of our hierarchical regression test we conducted to test
the moderating effect of workload on the relation between positive reciprocity and employee performance. The main effects are not significant; this implies that nor positive reciprocity (B
= -.04, p > .10) nor workload (B = -.05, p > .10) is related to employee performance.
However, the interaction effect shows marginally significance (B = .05, p < .10). Although, this moderation effect is positive, meaning that higher reciprocity results in less performance under a low workload, but not under a high workload (see the interaction plot in figure 2,
24 appendix). In hypothesis 5, we had formulated to expect a negative moderating effect.
Therefore, hypothesis 5 is rejected. This is an interesting finding, which we discuss later.
Concerning hypothesis 6, we have done a macro Hayes-analysis to test the complete model.
As a logical result of the rejection of hypotheses 3, 4 and 5, the complete model is not significant and hypothesis 6 has to be rejected as well.
Additional analysis
To investigate the positive moderation of workload , we have done some additional analyses.
Workload appears to be significant on the relation between LMX and employee performance as well. This moderation is also positive (B = .05, p <.10). Looking to both the correlations and the regressions, control variable ‘age’ seems to drive both interaction effects. When a regression is tested on the moderation effect of workload on the relationship between positive reciprocity and employee performance excluding age as control variable, there is no
significant effect (p-value of .26). Therefore, we tested whether there is a three-way
interaction with model 3 of Hayes (2013). It appeared that there is no three-way interaction effect (p-value of .80).
DISCUSSION Summary of results
In this study, we have examined the mediating role of positive reciprocity in the relationship between LMX and employee performance. Second, we have investigated the moderating effect of workload on the relationship between positive reciprocity and employee performance. Results of a sample of 366 employees and 53 supervisors show no support for these two expectations. In short, hypothesis 1 and 2 can be accepted: LMX is positively related to positive reciprocity as well as employee performance. Hypotheses 3, 4, 5 and 6 have been rejected: positive reciprocity appears not to be a mediator in the relationship between LMX and employee performance. Furthermore, workload does not moderate negatively on the
25 relationship between positive reciprocity and employee performance. Worth mentioning, is the result that workload does moderate on the relationship between positive reciprocity and employee performance. However, this moderating effect is positive.
Theoretical and managerial implications
Theoretical implications. In this paragraph linkages between our results and existing
literature will be shown. The hypotheses will be explained in subsequent order.
H1 & H2. As we named in the introduction, the positive relationship between LMX
and employee performance is widely supported. Examples of studies elaborating this
relationship are the studies of Gerstner & Day (1997), Graen & Uhl-Bien (1995) and Janssen
& Van Yperen, (2004). Our first hypothesis contributes to the robustness of the statement of this relationship. Our second hypothesis is of big value for research around the LMX theory.
As we mentioned in the theory section, we had reasons to expect the positive relation between LMX and positive reciprocity, mainly in the basics of the social exchange theory. Graen &
Uhl-Bien (1995) already suggested that positive reciprocity might be an important result of LMX relationships. However, this fundamental assumption had not yet been examined empirically. The results of this study show that positive reciprocal behaviour of an employee is a result of LMX. Therefore, this study fills a research gap by testing a fundamental
assumption of LMX theory. There remains, however, one point of discussion concerning this relationship. There are researchers mentioning (for example: Dienesch & Liden, 1986) that reciprocity is one of the foundations of LMX theory. This raises the question whether reciprocity is part of or result of LMX. Because LMX is about two-way exchanging, it is arguable that positive reciprocity might be necessary to maintain an LMX relationship. In distinguishing a mediator variable from a predictor variable Roe (2012) states that compiling more evidence is the thing to make it possible to know whether a construct can be seen as
26 mediator or part of the predictor. In other words, more empirical research is necessary to be able to state whether positive reciprocity is a result of LMX.
H3 & H4. Positive reciprocity was expected to be an explanation of the relationship
between LMX and an employee’s performance. This expectation was founded in three argued relations, namely at first the relation between LMX and employee performance, second, the relation between LMX and positive reciprocity and third, the relation between positive reciprocity and employee performance. The results show no support for this third relation, therefore we did not find evidence for a mediating effect which needs further elaboration.
We expected this relation for different studies show that positive reciprocity can operate as a motivator to do something extra (in this case: for the employee’s leader) (for example: Fehr & Gächter, 1998). In other words, as De Dreu & Nauta (2009) state, employees scoring high on positive reciprocity are willing to do something extra for their leader, by reciprocating the leaders’ behaviours. Our results might indicate that this willingness to do something extra has a different outcome than we expected. Namely, a different line of
reasoning is that people scoring high on positive reciprocity tend to reciprocate with extra-role behaviours, instead of in-role job performance. Examples of extra-role behaviours are: aiding fellow employees, taking actions that protect the organization from risk, offering constructive suggestions, and gaining knowledge and skills that are beneficial to the organization (George
& Brief, 1992). This thought is in line with the study of Settoon, Bennett & Liden (1996). In this study, they explain that an employee is doing more extra-role behaviours that benefit the supervisor like working overtime with the supervisor, or helping a co-worker with a task on which the supervisor is dependent when experiencing high quality LMX. So, we suggest that positive reciprocity is implemented mainly by extra-role behaviours, which are not measured by in-role job performance. This is an interesting thought which is worth measuring in future research. At all, in the search for mediating variables in the relationship between LMX and
27 employee performance, we now know that positive reciprocity is not the thing what explains this relationship.
H5 & H6. We found that a higher score on reciprocity results in less performance
under a low workload, but not under a high workload. This is a surprising finding. At first, we expected this moderating effect to be reversely. The negative moderation of workload was assumed, for previous research showed that a high workload is positively related to cognitive impairments (for example: difficulties in focusing and planning) (Boksem et al., 2005; Lorist et al., 2005; Van der Linden, 2006). These impairments might weaken the relation between positive reciprocity and performance. Another line of reasoning was the time and energy a high workload costs, whereupon an employee has its mind somewhere else (Firth et al, 2004).
Theoretically a high workload can lead to both psychological and physical demands for employees over sustained periods of time (Schaufeli, Bakker & van Rhenen, 2009). We expected that these demands would weaken the relation between positive reciprocity and performance.
It might be that workload does not affect employee performance of employees scoring high on LMX and thus high on positive reciprocity, for these employees can better cope with a high workload (Bakker et al., 2003; Liden & Graen, 1980). An explanation of these better coping skills is that employees having a high quality LMX relationship are provided with more resources, like autonomy, possibilities for professional development and performance feedback (Bakker et al., 2003). So, workload would not affect employees scoring high on positive reciprocity at all, because of their coping skills. This would put the danger of a too high workload in a new perspective, for results of this study show for these strong reciprocal employees no difference in performance between employees having a high or low workload.
However, a difference in performance is visible for employees scoring low on positive reciprocity. In this differentiation, employees having a low workload perform better than
28 employees having a high workload. For these employees (scoring low on positive
reciprocity), it seems to appear that a high workload lowers their performance. Then, our argumentation about the negative influence of a high workload might be the case. This
argumentation included that workload is positively related to mental depletion and stress (Van der Hulst & Geurts, 2001; Boksem et al., 2005; Lorist et al., 2005). On its turn, mental
depletion and stress are related with reduced cognitive and behavioural performance (Boksem et al., 2005; Lorist et al., 2005). As a result, we expected workload to weaken the relationship between positive reciprocity and employee performance.
Although, the employees with low reciprocity and a low workload perform also better than the employees scoring high on positive reciprocity. In fact, it appears that these
employees showing few reciprocal behaviours, with a low workload, perform best. This interesting conclusion might indicate that employees scoring high on positive reciprocity are scoring lower on their own performance, because they are spending time and energy in the
‘performance’ of others. This refers to the difference between extra-role behaviours and in- role job performance we named; that employees with low reciprocity are more focused on their own, routine, tasks (Graen & Uhl-Bien, 1995), where employees with high reciprocity tend to exhibit more extra-role behaviours (Graen & Uhl-Bien, 1995) which are not measured by in-role job performance. So, an explanation of the positive moderation as shown in the interaction plot in figure 2, lays in the better coping skills (for high reciprocators), the
cognitive impairments a high workload costs (for low reciprocators) and the time and energy extra-role behaviours cost (the fact that employees with low positive reciprocity and low workload score best).
We have tried to answer the question which has been a research gap for a long time:
why LMX leads to employee performance. Unfortunately, this study cannot provide a direct answer to this question. However, the insignificance of positive reciprocity as mediator is an
29 important finding. Namely, as positive reciprocity is an important outcome of LMX, we now know that positive reciprocity is not the reason why LMX leads to employee performance. By conducting studies like this, it will become clear in the end what explains the relation between LMX and employee performance.
The positive moderation of workload on the relation between positive reciprocity and employee performance is completely new and is, therefore, of added value to the existing research. This provides insight in the facilitating conditions under which a reciprocating employee performs. The concrete influence we found, also gives a relativistic view on
workload. Namely, we have support to state that under certain conditions a different workload is requisite. Concretely, a high workload does not always affect performance. This finding contradicts earlier research of for example Van der Linden (2006); studies which only name negative consequences of a high workload. With the nowadays increasing complaints about mental depletion (which is seen as a result of a too high workload, by for example Boksem &
Tops, 2008; Ricci et al., 2007), we have given insight in situations when workload does not affect employee performance. Therefore, we have given a start to research around the phenomenon ‘situational workload’. We come back to this term under ‘future directions’.
Managerial implications. Despite the fact that our research has not produced the expected
outcomes, our study has strong managerial implications. Our first advice for managers is to invest in an LMX relationship. Supported by existing research, this study shows that a high LMX relationship increases employee performance. Our advice implicates then to develop two-way exchanging with each of the employees personally. As Graen & Uhl-Bien (1995) describe, an LMX relationship can be increased by providing the employee special privileges (like access to key information), career-enhancing opportunities (like special work
assignments) and a higher discretion in doing their jobs. Besides, make sure that you create and maintain an atmosphere of trust and respect, at first by leading by example.
30 Our second advice is about workload. As we have seen, workload has a significant influence on employee performance, on both the relation between LMX and employee performance as well as the relation between positive reciprocity and employee performance.
Therefore, it is advisable for managers to consider the workload and performance
measurement of employees. This study can not provide a deep insight in the question why the moderation-effect is positive. Although, in the case that an employee experiences a low workload, it might be that an employee scoring high on positive reciprocity is scoring lower on its own performance due to the fact that they are spending time and energy in the
‘performance’ of others. Therefore, we advise managers to consider the workload and performance of an employee more personally, mainly during performance appraisals. It is valuable knowledge to have a good insight in an employee’s contribution to a company.
Therefore, we recommend to evaluate the way performance of an employee is measured: to make sure that reciprocal behaviours will not be unseen by a manager.
Limitations and future directions
Limitations. Besides our different contributions, we can also designate some limitations. In
this study, data has been used from very different types of companies. This may be a good point regarding the generalizability of this study. However, two footnotes have to be made. At first, from the viewpoint of situational leadership, it is arguable whether it is wise to include such a variety in types of companies. Namely, as Yukl (2013) elaborates the theory of situational leadership, we may know that in different situations, different ways of leadership are appropriate. Then, it is possible that certain results are distorted. For example: from the viewpoints of a supervisor working for the police as well as the viewpoint of a supervisor within a school, ‘relation’ and ‘trust’ may be very different constructs. Second, our results show a large majority of the quartiaire (non-commercial) sector: 33 of 53 companies. This
31 sector is out of the four sectors overrepresented and this may weaken the generalizability. It might be better to focus on one sector, or to focus more on an equal distribution of sectors.
Considering our method, we have to recommendations for future research. At first, this study only conducts quantitative research. As we have experienced, in facing unexpected results, it is hard to answer why-questions with only data and existing literature. Therefore, we recommend for future research to add a qualitative element to the existing way of quantitative measurement in the form of interviews with supervisors and employees.
Secondly, in assessing their employees or leaders, the respondents were potentially biased; it might be that a respondent will experience a kind of social pressure to score certain answers higher or lower, also when he or she knows that the scores are anonymous. Therefore, it is considerable to add a third party in assessing leaders and subordinates, for third parties are in most cases more objective. This might add validity to the test results.
Future directions. An important goal was to give insight in the positive relation between
LMX and employee performance. We didn’t find positive reciprocity to be the mediating factor in this relation, therefore, we recommend to continue the examination of the positive relation between LMX and employee performance, with testing other mediators.
Besides, as named earlier, the positive moderation of workload raises questions.
Future research should mainly unravel the exact influence of workload on the relation between LMX and employee performance. We recommend to take more differentiations of workload into account, for this is a broad construct. For example, workload can be examined quantitatively or qualitatively, which may have quite different results. Furthermore, we advise to investigate more concerning the influence of workload. As we argued, as in line with the basics of ‘situational leadership’ (Yukl, 2013), by investigating more in this topic, we can create a new construct named ‘situational workload’. The idea of this construct is, as is the case with situational leadership, to elaborate when it is necessary to take the workload of an
32 employee into account in order to enhance his or her performance. As we argued, it could be that a workload does not affect the performance of a high reciprocating employee because of he or she has better coping skills. These coping skills could arise from the fact that employees scoring high on LMX and thus high on positive reciprocity receive more resources from their leaders. This thought has to be examined empirically. If this is the case, we get insight in a way to eliminate the negative consequences a high workload can have, namely by giving more resources to employees.
CONCLUSION
As pioneer in economic research, moral philosopher and economist Adam Smith left a famous quote: "Every man lives by exchanging". This study again endorses the important role of exchanging between people. This importance appears in the higher in-role job performance of employees with a high exchanging (LMX) relationship. In addition, we have tested a
fundamental assumption of LMX theory; an examination of the ‘how’ of the relationship between LMX and employee performance. It appears that positive reciprocity is not the thing why exchanging between leader and employee leads to employee performance. Besides addressing the knowledge gap concerning the ‘how’, we have also examined a ‘when’- variable. Surprisingly, low reciprocators with a low workload perform best. The moderation of workload has raised a lot of questions; this opens a number of research opportunities. We hope these results will stimulate future research to continue further examination of mediators and moderators on the LMX-employee performance relationship.
33 References
Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and interpreting interactions. Sage.
Bakker, A. B., Demerouti, E., Taris, T. W., Schaufeli, W. B., & Schreurs, P. J. (2003).
multigroup analysis of the job demands-resources model in four home care organizations. International Journal of stress management, 10(1), 16.
Bauer, T. N., & Green, S. G. (1996). Development of Leader-Member exchange: a longitudinal test. Academy of management journal, 39(6), 1538-1567.
Boksem, M. A. S., Lorist, M. M., & Meijman, T. F. (2005). Effects of mental fatigue on attention: an ERP study. Cognitive Brain Resource, 25(1), 107-116.
Boksem, M. A. S., Meijman, T. F., & Lorist, M. M. (2006). Mental fatigue, motivation and action monitoring. Biological Psychology, 72(2), 123-132.
Boksem, M. A. S., & Tops, M. (2008). Mental fatigue: costs and benefits. Brain research reviews, 59(1), 125-139.
Borman, W. C., & Motowidlo, S. J. (1997). Task performance and contextual performance:
The meaning for personnel selection research. Human Performance, 10(2), 99-109.
Centraal Bureau voor Statistiek. (2016) Ontleend van:
http://statline.cbs.nl/statweb/publication/?vw=t&dm=slnl&pa=81589ned&d1=0&d2=
,2,50,68,470,484,504,549,847,910,936,986,1043,1057,1133,1209,1225,1259,1318,13 2,1437,1446&d3=(l-15)-l&hd=160224-1421&hdr=t,g1&stb=g2
Dansereau, E., Cashman, J., & Graen, G. (1973). Instrumentality theory and equity theory as complementary approaches in predicting the relationship of leadership and turnover among managers. Organizational Behavior and Human Performance, 10(2), 184-200.
34 Dansereau, E., Graen, G., & Haga, W. J. (1975). A vertical dyad linkage approach to
leadership within formal organizations: A longitudinal investigation of the role making process. Organizational Behavior and Human Performance, 13(1), 46-78.
Dasborough, M. T., & Ashkanasy, N. M. (2002). Emotion and attribution of intentionality in leader-member relationships. Leadership Quarterly, 13(5), 615-634.
Dienesch, R. M., & Liden, R. C. (1986). Leader-Member Exchange model of leadership: a critique and further development. The academy of management review, 11(3), 818-634.
De Dios Tena, J., & Forrest, D. (2007). Within-season dismissal of football coaches:
statistical analysis of causes and consequences. European journal of operational research, 181(1), 362-373.
Dockery, T. M., & Steiner, D. D. (1990). The role of the initial interaction in leader-member exchange. Group and organization studies, 15(4), 395-413.
Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of
Management Journal, 45(4), 735-744.
Fehr, E., & Gächter, S. (1998). Reciprocity and economics: the economic implications of Homo Reciprocans. European economic review, 42(3), 845-859.
Fehr, E., & Gächter, S. (2000). Fairness and retaliation: the economics of reciprocity. The journal of economic perspectives, 14(3), 159-181.
Fehr, E., Kirchsteiger, G., & Riedl, A. (1993). Does fairness prevent market clearing? An experimental investigation. Quarterly Journal of Economics 108(2), 437-459.
35 Gerstner, C. R., & Day, D. V. (1997). Meta-Analytic review of leader–member exchange
theory: Correlates and construct issues. Journal of applied psychology, 82(6), 827-844.
Foster, R., & Kaplan, S. (2001). Creative destruction. Why companies that are built to last under-perform the market – and how to successfully transform them. New York:
Currency.
George, J. M., & Brief, A. P. (1992). Feeling good-doing good: A conceptual analysis of the mood at work-organizational spontaneity relationship. Psychological Bulletin, 112(2), 310.
Graen, G., & Cashman, J. F. (1975). A role making model in formal organizations: A developmental approach, Leadership Frontiers, 143-165.
Graen, G.B., Novak, M., & Sommerkamp, P. (1982). The effects of leader-member exchange and job design on productivity and satisfaction: Testing a dual attachment model.
Organizational Behavior and Human Performance, 30(1), 109-131.
Graen, G. B., & Scandura, T. A. (1987). Toward a psychology of dyadic organizing, Research in organizational behaviour, 9, 175-208.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership:
Development of leader-member exchange (LMX) theory of leadership over 25 years:
Applying a multi-level multi-domain perspective. The leadership quarterly, 6(2), 219-247.
Hart, S. G., & Staveland, L. E. (1988). Development of NASA-TLX(Task Load Index):
Results of empirical and theoretical research. Advances in Psychology, 52, 139-183.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press.
36 Hockey, G. R. J. (1997). Compensatory control in the regulation of human performance under
stress and high workload: a cognitive-energetical framework. Biological Psychology, 45(1), 73-93.
Janssen, O., & Van Yperen, N. W. (2004). Employees’ goal orientations, the quality of leader-member exchange, and the outcomes of job performance and job satisfaction.
Academy of management journal, 47(3), 368-384.
Jex, S. M. (1998). Stress and job performance: Theory, research, and implications for managerial practice. Thousand Oaks, CA: Sage
Karsan, R. (2007). Calcaluting the cost of turnover, Employment relations today,34(1), 33-36.
doi: 10.1002/ert.20139
Kuoppala, J., Lamminpaa, A., Liira, J., & Vainio, H. (2008). Leadership, job well-being, and health effects: A systematic review and meta-analysis. Journal of Occupational and Environmental Medicine, 50(8), 904-915. doi: 10.1097/JOM.0b013e31817e918
Lewis, J. D., & Weigert, A. (1985). Trust as social reality. Social Forces, 63(4), 967-985.
Liden, R. C., & Graen, G. (1980). Generalizibility of the vertical dyad linkage model of leadership. Academy of management journal, 23(3), 451-465.
Liden. R. C., & Maslyn, J. M. (1998). Multidimensionality of leader-member exchange: an empirical assessment through scale development. Journal of management, 24(1), 43-72.
Liden, R. C., Sparrowe, R. T., & Wayne S. J. (1997). Leader-member exchange theory: the past and potential for the future. Research in personnel and human resources management, 15, 47-120.
37 Liden, R. C., Wayne, S. J., & Stilwell, D. (1993). A longitudinal study on the early
development of leader-member exchanges. Journal of applied psychology, 78(4), 662.
Lorist, M. M., Boksem, M. A. S., & Ridderinkhof, K. R. (2005). Impaired cognitive control and reduced cingulate activity during mental fatigue, Cognitive Brain Resources, 24(2), 199-205.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709-734.
McCabe, K. A., Rigdon, M. L., & Smith, V. L. (2003). Positive reciprocity and intentions in trust games. Journal of Economic Behavior & Organization, 52(2), 267-275.
Oldham, G. R., Cummings, R. (1996). Employee creativity: personal and contextual factors at work. Academy of management journal, 39(3), 607-634.
Perugini, M., Gallucci, M., Presaghi, F., & Ercolani, A. P. (2003). The personal norm of reciprocity. European Journal of Personality, 17(4), 251-283. doi: 10.1002/per.474 Ricci, J. A., Chee, E., Lorandeau, A. L., & Berger, J. (2007). Fatigue in the US workforce:
prevalence and implications for lost productive work time. Journal of occupational environmental medics, 49(1), 1-10.
Roe, R. A. (2012). What is wrong with mediators and moderators. The European Health
Psychologist, 14(1), 4-10.
Schaufeli, W. B., Bakker, A. B., & Van Rhenen, W. (2009). How changes in job demands and resources predict burnout, work engagement, and sickness absenteeism. Journal of Organizational Behavior, 30(7), 893-917. doi: 10.1002/job.595
