Journal of Managerial Psychology

Leader-member exchange, work engagement, and job performance

Kimberley Breevaart Arnold B. Bakker Evangelia Demerouti Machteld van den Heuvel
Article information:
To cite this document:
Kimberley Breevaart Arnold B. Bakker Evangelia Demerouti Machteld van den Heuvel , (2015),"Leader-member exchange,
work engagement, and job performance", Journal of Managerial Psychology, Vol. 30 Iss 7 pp. -
Permanent link to this document:
http://dx.doi.org/10.1108/JMP-03-2013-0088
Downloaded on: 08 August 2015, At: 06:59 (PT)
References: this document contains references to 0 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 39 times since 2015*
Users who downloaded this article also downloaded:
Elaine Farndale, Inge Murrer, (2015),"Job resources and employee engagement: a cross-national study", Journal of
Managerial Psychology, Vol. 30 Iss 5 pp. 610-626 http://dx.doi.org/10.1108/JMP-09-2013-0318
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Kurt Matzler, Florian Bauer, Todd A Mooradian, (2015),"Self-esteem and transformational leadership", Journal of Managerial
Psychology, Vol. 30 Iss 7 pp. -
Alexandre Bachkirov, (2015),"Managerial decision making under specific emotions", Journal of Managerial Psychology, Vol.
30 Iss 7 pp. -

Access to this document was granted through an Emerald subscription provided by emerald-srm:463575 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 LMX AND JOB PERFORMANCE 1

Leader-Member Exchange, Work Engagement, and Job Performance

Leader-member exchange theory (LMX theory; Graen & Cashman, 1975; Graen & Uhl-

Bien, 1995) is unique in its focus on the dyadic relationship between leader and follower. Rooted

in role-making and social exchange theories (Blau, 1964; Graen, 1976; Kahn, Wolfe, Quinn,

Snoek, & Rosenthal, 1964), LMX theory states that followers develop unique exchange

relationships with their leader. In turn, the quality of this relationship influences followers’ work

attitudes and behaviors. Consistent with these ideas, meta-analytic studies show that the quality
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

of the LMX relationship is related to a range of positive follower outcomes, like job satisfaction,

task performance, organizational citizenship behavior (OCB), commitment, and role clarity (e.g.,

Dulebohn, Bommer, Liden, Brouer, & Ferris, 2012; Ilies, Nahrgang, & Morgeson, 2007;

Volmer, Niessen, Spurk, Linz, & Abele, 2011). However, although there is a wealth of literature

on the proximal effects of LMX on follower outcomes (e.g., Dulebohn et al., 2012; Gerstner &

Day, 1997), little is known about the process through which leader-member exchanges influence

follower outcomes.

The current study contributes to the LMX literature by examining LMX as a distal

predictor of followers’ job performance. We are among the first to study the process underlying

the relationship between LMX and followers’ job performance and to our knowledge, the first to

examine the relationship between LMX and work engagement. Based on conservation of

resources theory (Hobfoll, 1989; 2001) and job demands-resources theory (Bakker & Demerouti,

2014; Demerouti, Bakker, Schaufeli, & Nachreiner, 2001), we argue that LMX is positively

related to followers’ job performance, because followers have access to more job resources when

they have a high-quality relationship with their leader and are therefore more engaged in their
 LMX AND JOB PERFORMANCE 2

work. We examine these relationships within the hierarchical structure of the Dutch police force,

where leadership plays a pronounced role.

LMX and Follower Job Performance

LMX theory proposes that leaders have unique social exchange relationships with their

followers and that the quality of these relationships (ranging from low to high) differs between

employees with the same leader (Graen & Uhl-Bien, 1995; Liden, Sparrowe, & Wayne, 1997).

Low-quality LMX relationships are based on economic exchanges, i.e., exchanges based on the
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

formal requirements of the employment contract in which employees do what they are expected

to do and get paid accordingly. In contrast, high-quality exchanges go beyond the formal contract

and are based on trust, mutual obligation, and mutual respect and result in affective attachment.

The type of LMX relationship that develops depends on the result of a series of role making

episodes in which leaders express their expectations and employees show the degree to which

they are able and willing to live up to these expectations.

The quality of the LMX relationship determines the degree to which leaders reciprocate

meeting certain job demands by employees with additional resources like autonomy,

information, and the opportunity to participate in the decision making process. Graen and

Cashman (1975) argue that these additional resources explain why the quality of the LMX

relationship contributes to employees’ job performance. Put differently, high LMX relationships

are characterized by high expectations regarding employees’ performance, in return for the

investments made by the leader. Research confirms that members in high-quality LMX

relationships perform better. In their meta-analyses, Gerstner and Day (1997) and Dulebohn et al.

(2012) showed that LMX is positively related to both subjective and objective performance. The

question that we will answer with this study is why this is the case.
 LMX AND JOB PERFORMANCE 3

LMX, Work Engagement, and Job Performance

We argue that LMX is positively related to followers’ job performance, because high-

quality LMX relationships enhance followers’ work engagement. Work engagement is a

positive, work-related state of mind that is characterized by vigor, dedication, and absorption

(Schaufeli, Bakker & Salanova, 2006). Thus, engaged employees have high levels of energy, are

enthusiastic about, inspired by and proud of their work, and feel like time flies when they are

working. In the current economic situation, having an engaged workforce may provide a
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

competitive advantage, because work engagement is an active state that is positively related to

important outcomes such as job performance, commitment and health (for meta-analyses see

Christian, Garza, & Slaughter, 2011; Halbesleben, 2010).

According to conservation of resources theory (COR theory; Hobfoll, 1989; 2001),

people are motivated to obtain, retain, protect and foster their resources (e.g., autonomy,

developmental opportunities, social support). Leaders, in their inherent position of power, are an

important source of support and research has shown that social support is positively related to

work engagement (Halbesleben, 2010). According to the Job Demands-Resources model,

employees are especially engaged in their work when their resources are combined with

challenging demands (JD-R model; Bakker & Demerouti, 2007; 2014; Demerouti et al., 2001).

Accordingly, it is likely that employees feel more engaged when they have a high-quality

exchange relationship, because their leader facilitates their job performance, but also expects

high job performance in return.

From a social exchange perspective, high-quality LMX relationships may contribute to

employees’ intrinsic motivation to do their job well, making it likely that employees in high-

quality LMX relationships become engaged in their work. It has been shown that supervisors in
 LMX AND JOB PERFORMANCE 4

high-quality LMX relationships give their followers more intrinsic (empowerment, praise) and

extrinsic (salary raise) rewards, which result in more positive attitudes toward work (Epitropaki

& Martin, 2005). Finally, followers in a high-quality relationship have been found to be

optimistic and self-efficacious (Vasudevan, 1993), and such self-beliefs are important predictors

of work engagement (Halbesleben, 2010). Therefore, we hypothesize:

Hypothesis 1: Work engagement mediates the relationship between LMX and job performance.

LMX, Job Resources, Work Engagement, and Job Performance

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

The assumption that LMX is related to follower outcomes because leaders form a

resourceful work environment is in line with some findings that leaders in high-quality LMX

relationships provide employees with decision making latitude, empowerment, and social

support (e.g., Keller & Dansereau, 1995; Scandura, Graen, & Novak, 1986; Sparrowe & Liden,

1997). However, when relating LMX to job-related outcome variables, the provision of job

resources is often assumed, but not measured. Since the exchange of resources is a central

feature of LMX theory, in the current study, we explicitly measure followers’ job resources to

examine whether they can explain the relationship between LMX and follower job outcomes. We

focus on three of the most often studied job resources; autonomy, developmental opportunities,

and social support from coworkers (Halbesleben, 2010).

Leaders’ investment in high-quality LMX relationships creates positive expectations

regarding employees’ job performance. LMX theory posits that leaders’ self-image are damaged

when these expectations are not met and therefore these leaders often facilitate high

performance. Research has indeed shown that leaders in high-quality LMX relationships reduce

role conflict, role ambiguity and role overload (e.g., Dunegan, Uhl-Bien, & Duchon, 2002;

Lagace, Castleberry, & Ridnour, 1993). Besides, since employees in high-quality relationships
 LMX AND JOB PERFORMANCE 5

are trusted by their leader, they are provided with more decision latitude (Townsend, Da Silva,

Mueller, Curtin, & Tetrick, 2002) and empowerment (e.g., Keller & Dansereau, 1995). This

provides employees in high-quality LMX relationships the freedom to decide for themselves

which work assignments they will focus on, and how they will execute them. Based on these

arguments, we expect LMX to be positively related to autonomy.

Next, we expect employees in high-quality relationships to have more developmental

opportunities compared to their counterparts. For example, employees in high-quality LMX

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

relationships have a privileged way of communication with the leader and are provided with

desirable work assignments, while employees in low-quality relationships rarely meet with their

supervisors and are often provided with undesirable monotonous assignments (Dulebohn et al.,

2012). This means that particularly employees in high-quality relationships are able to work on

their self-growth. These employees thereby become even more valuable to the leader and

maintain the quality of the relationship with their leader. This relationship has also been

described as a mentoring relationship (Scandura & Williams, 2004), in which the leader acts as a

coach and invests in the career success of the employee (Sosik & Godshalk, 2000).

Finally, we expect LMX to be positively related to social support from coworkers, since

relationships in one part of the organization may influence relationships in other parts of the

organization (Graen & Uhl-Bien, 1995). Research indeed shows that the quality of the LMX

relationship with the leader influences the relationships between coworkers (Sherony & Green,

2002). More specifically, employees in a high-quality relationship with their supervisor had

significantly higher quality exchange relationships with coworkers who were also in a high-

quality LMX relationship with the same supervisor. In this case, both coworkers share the same

positive experiences, so they are in a similar situation (Heider, 1958; Sherony & Green, 2002).
 LMX AND JOB PERFORMANCE 6

Also, Ilies et al. (2007) showed in their meta-analysis that LMX quality is positively related to

organizational citizenship behavior (OCB). This means that employees in high-quality LMX

relationships engage in behavior that is not defined in their role description, like helping

colleagues with a high workload or helping employees who have been absent. These helping

behaviors may create a work environment in which colleagues help and support each other.

Based on these arguments and earlier research on the relationship between job resources and

work engagement (Xanthopoulou et al., 2008, 2009), we hypothesize:

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Hypothesis 2: The relationship between LMX and job performance is sequentially mediated by

job resources (autonomy, developmental opportunities, social support), and work engagement

(all relationships are positive).

Figure 1 provides an overview of all hypothesized relationships.

Method

Participants and Procedure

Participants were Dutch police officers working within one district of the Dutch police

force. After general communications about the research, the invitation to participate in an on-line

survey was sent out to all 1780 employees via email. A total of 950 police officers completed the

survey (response rate = 53%). The questionnaires were filled in anonymously, but participants

were asked to indicate to which team they belonged by selecting their team from a list.

Employees were asked to fill out the LMX questions while keeping in mind one specific leader.

Finally, this resulted in 847 participating employees from 58 teams. Participants could request a

personalized feedback report on their responses.

The sample consisted of 527 male employees (62.2%) and 320 female employees

(37.8%). The mean age of the participants was 41.9 years (SD = 10.5) and mean organizational
 LMX AND JOB PERFORMANCE 7

tenure was 16.3 years (SD = 11.41). The majority of the participants was either married,

cohabiting or had a steady relationship (89.3%) and 72.2% worked 36 hours or more per week.

The mean number of team members in each team was 25.8, meaning that teams had 26 members

on average.

Measures

Control variables. We measured and included several demographic (i.e., gender, age,

education, and marital status) and work-related (i.e., working hours and tenure) background
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

variables.

Leader-member exchange was measured using the Dutch version (see LeBlanc, 1994) of

the Leader-Member Exchange scale (Graen & Uhl-Bien, 1995). This scale consists of five items

rated on a 5-point scale (1 = never, 5 = often). An example item is: ‘My supervisor uses his/her

influence to help me with problems at work’. The internal consistency of this scale was high

(Cronbach’s α = .91).

Job resources were measured with items developed by Bakker, Demerouti, Taris,

Schaufeli, and Schreurs (2003). All items were measured on a 5-point scale (1 = never, 5 =

often). An example item of each job resource is ‘I am able to decide myself how to execute my

work’ (autonomy), ‘My work offers me the opportunity to learn new things’ (developmental

opportunities), and ‘When it is necessary, I can ask my colleagues for help’ (social support).

Resources were measured with four items each, except for developmental opportunities, which

was measured with three items. Internal consistencies of the scales were .81 for autonomy, .87

for social support and .89 for developmental opportunities.

Work engagement was measured using the nine-item version of the Utrecht Work

Engagement Scale (Schaufeli, Bakker, & Salanova, 2006). Work engagement consists of three
 LMX AND JOB PERFORMANCE 8

dimensions (vigor, dedication, and absorption) that were measured with three items each.

Example items are: ‘At work, I feel bursting with energy’ (vigor), ‘I am enthusiastic about my

work’ (dedication), and ‘I am immersed in my work’ (absorption), which had to be answered on

a 6-point scale (0 = never, 6 = always). The internal consistency of this scale was high (α = .95).

Job performance was measured with three items from Goodman and Svyantek (1999) to

measure task performance. The validity of the selected items was supported by Xanthopoulou et

al. (2008). An example item is: ‘I perform well on the core aspects of my work’. The items were
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

answered on a 6-point scale (1 = totally disagree, 6 = totally agree). The internal consistency of

this scale was good (α = .86).

Strategy of Analysis

The individuals in our sample were nested within teams, thereby violating the

independence assumption underlying many statistical techniques. To account for the nested

structure of the data, we used multilevel structural equation modeling (MSEM) using Mplus

(Muthén & Muthén, 1998-2010). We have a two-level model with individuals at the first level

(Level 1; N = 527), and teams at the second level (Level 2; N = 58). We followed Maas and

Hox’s (2005) rule of thumb for power in multilevel modeling. This rule states that a minimum of

30 cases at the highest level is required for robust estimations.

The use of multilevel analyses is justified when there is sufficient variability at both

levels of analysis. The ICC’s indicated that the variance explained by the team level ranged from

2.7% for job performance to 17.6% in autonomy. When multilevel data are analyzed on a single

level, parameter estimates can be affected which may result in inaccurate statistical inferences.

Since we were only interested in the first (individual) level, we used multilevel analyses to

control for the possible confounding influence of variance at the second (team) on our results. As
 LMX AND JOB PERFORMANCE 9

alluded to above, we used multilevel analyses because regular structural equation modeling

analyses would violate the independence assumption underlying this technique (Hox, 2010).

Results

Descriptive Statistics

Table 1 shows the means, standard deviations, inter-correlations, and internal

consistencies of the study variables.

Measurement Model
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

First, we tested the measurement model to examine the construct validity of our

variables. The measurement model consisted of the study variables with scale items reflecting

their respective latent construct. Specifically, the measurement model consisted of six factors,

including LMX (five items), autonomy (four items), developmental opportunities (three items),

social support (four items), work engagement (nine items), and job performance (three items)

with scale items tapping the latent construct. This measurement model showed good fit to the

data (CFI = .93; TLI = .92; RMSEA = .06; SRMR = .04), and all items had significant loadings

on the intended latent factors (.56 - .89, p < .001). Next, we compared this measurement model

to a one-factor, four-factor (i.e., all job resources combined into one factor) and a five-factor

(LMX and social support as one factor ) model (see Table 2) and found that the proposed

measurement model fitted best to the data.

Structural and Alternative Models

Next, we tested our structural models using multilevel structural equation modeling (see

Table 3). In all analyses, we controlled for gender, age, marital status, education, working hours

per week, and tenure, because they were related to our study variables. To test the significance of

the mediation effects, we used the online interactive tool developed by Selig and Preacher
 LMX AND JOB PERFORMANCE 10

(2008). This tool uses the parametric bootstrap method to create confidence intervals without

making any assumptions about the distribution of the indirect effect. Hypothesis 1 states that the

relationship between LMX and job performance is mediated by work engagement. The path from

LMX to work engagement was .46 (p < .001, 95% CI [.41, .51]) and the path from work

engagement to job performance was .34 (p < .001, 95% CI [.26, .41]). Furthermore, there was a

significant mediation effect (.15, p < .001, 95% CI [.12, .20]). This model fit well to the data

(CFI = .91, RMSEA = .07, SRMR = .03).

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

We compared our hypothesized model to the partially mediated model (i.e., including the

direct effect from LMX to job performance), but there was no significant decrease in χ2 (∆χ2 (1)

= .01, n.s.). Therefore, we prefer our hypothesized, more parsimonious model. Next, we

compared our model to the direct effects only model, including paths from LMX and work

engagement to job performance. We compared the fit of our hypothesized model to the fit of the

direct effects only model, which showed a significant increase in χ2 (∆χ2 (5) = 31.11, p < .001),

meaning that our hypothesized model fits better to the data.

Hypothesis 2 states that the relationship between LMX and job performance is

sequentially mediated by job resources (autonomy, developmental opportunities, and social

support) and work engagement. The results show that LMX was positively related to autonomy

(.40, p < .001, 95% CI [.35, .45]), social support (.39, p < .001, 95% CI [.34, .45]), and

developmental opportunities (.51, p < .001, 95% CI [.47, .56]). In turn, autonomy (.12, p < .05,

95% CI [.03, .20]), social support (.29, p < .001, 95% CI [.24, .34]) and developmental

opportunities (.41, p < .001, 95% CI [.33, .49]) were positively related to work engagement.

Finally, work engagement was positively related to job performance (.34, p < .001, 95% CI [.26,

.41]). The results of the structural model supported Hypothesis 2 for autonomy (.01, p < .05,
 LMX AND JOB PERFORMANCE 11

95% CI [.002, .02]), developmental opportunities (.04, p < .001, 95% CI [.02, .05]) and social

support (.02, p < .001, 95% CI [.01, .03]). This model showed a satisfactory fit to the data (CFI =

.92, RMSEA = .06, SRMR = .04; Hoyle, 1995, Kline, 2005; MacCallum, Browne, & Sugawara,

1996). We compared our hypothesized model to a model including the direct paths from LMX to

work engagement. There was a significant decrease in χ2 (∆χ2 (1) = 17.49, p < .001) and

therefore we prefer the partially mediated model. Next, we added the direct paths from all job

resources to job performance, but this did not result in a better model fit (∆χ2 (3) = 1.5, n.s.).
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Finally, we compared our hypothesized model to the direct effects only model, including paths

from LMX, job resources and work engagement to job performance. This comparison showed a

significant decrease in χ2 (∆χ2 (5) = 15.52, p < .01), indicating that our hypothesized model fits

better to the data.

We used contrast effects to test the relative importance of the job resources. The contrasts

indicated that social support (-.02, p < .01, 95% CI [-.03, -.01]) and developmental opportunities

(-.03, p < .01, 95% CI [.-05, -.02]) were more important mediators compared to autonomy.

Besides, there was a significant difference between developmental opportunities and social

support as mediators (-.01, p < .01, 95% CI [-.02, -.003]). That is, developmental opportunities is

a more important mediator compared to social support. Taken together, these results provide

partial support for Hypothesis 2b. The final model as displayed in Figure 2 explains 25.6% of the

variance in autonomy, 28.8% of the variance in developmental opportunities, 20.1% in social

support, 39.8% in work engagement, and 13.3% in job performance. The figure shows the

standardized estimates of all the paths in the final model. All estimates are significant at p < .001,

except the estimate of the autonomy-work engagement relationship, which is nonsignificant.

Discussion
 LMX AND JOB PERFORMANCE 12

This study is one of the first to examine LMX as a distal predictor of job performance

and relatedly, one of the first to test a sequentially mediating mechanism that can account for the

LMX-job performance relationship. In addition, this study is innovative in that it is one of the

first to test a sequential mediation model using structural equation modeling. Furthermore, to our

knowledge, this is the first study that examines the relationship between LMX and work

engagement. Finally, our sample consisted of a large number of police officers, for whom

leadership is very relevant part of everyday work life. The results largely confirm our hypotheses
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

by showing that high-quality LMX relationships initiate a motivational process, whereby the

relationship between LMX and subordinates’ job performance is sequentially mediated by

employees’ job resources (developmental opportunities and social support) and work

engagement.

Job Resources as Mediators

This study contributes to the literature on LMX theory by showing that leaders can foster

the availability of job resources, which enhances employees’ work engagement and job

performance. In line with COR theory, LMX proved to be an important resource from which

other resources can be build (i.e., autonomy, developmental opportunities and social support).

Although it has been shown that LMX is directly and positively related to job performance (e.g.,

Gerstner & Day, 1997) and to job resources (e.g., Sparrowe & Liden, 1997) not much is known

about how LMX and job performance are related. Our study suggests that leaders can positively

influence their followers’ work engagement, both directly by the effect of the quality of their

relationship and indirectly through their influence on the availability of job resources to their

employees (mainly through developmental opportunities). The latter may be especially

 LMX AND JOB PERFORMANCE 13

interesting when employees have difficulties creating their own resources, caused by very strict

rules or the individuals’ lack of proactive behavior.

In the past, LMX has been considered as a type of coaching from the leader within the

JD-R model (e.g., Xanthopoulou et al., 2009). Although LMX can be considered as a job

resource, post hoc analyses showed that the model with LMX as an antecedent of other job

resources fit the data better than the model with LMX as a job resource not preceding other

resources. This underscores the role of the supervisor in creating resourceful work environments
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

for their subordinates.

Job Resources and Work Engagement as Sequential Mediators

Having a high-quality relationship not only contributes to employees’ work engagement,

but indirectly also positively influences the organization at large. This is because the quality of

the LMX relationship is positively related to employees’ job performance and stimulates the

initiation of a motivational process (i.e., the provision of job resources that are positively related

to work engagement). This contributes to the LMX literature by showing that there are important

intervening processes that account for the LMX-job performance relationship (Dulebon et al.,

2011; Gerstner & Day, 1997) and by showing that LMX is also a proximal predictor of employee

well-being. In this study, the relationship between LMX and job performance is even fully

mediated by job resources and work engagement, suggesting that followers job performance is a

more distal consequence of LMX.

Autonomy appeared to be the least strong mediator compared to social support and

developmental opportunities. An explanation could be that autonomy may be less important for

employees within the police force than for other less ‘protocolized’ occupational groups. This is

in line with the JD-R model, which assumes that each profession has its own unique combination
 LMX AND JOB PERFORMANCE 14

of job resources and job demands. In the police force, there are strict rules and protocols to be

adhered to. Subordinates may be used to these rules, which could explain why their engagement

is less dependent on the amount of autonomy they have within their job.

Practical Implications

The abovementioned results emphasize the importance for subordinates to have a good

relationship with their leader, since the quality of the LMX relationship is associated with the

quality of the work environment. It also stresses the importance for leaders of having a good
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

relationship with subordinates, since this is positively related to employees’ work engagement

and their appraisals of job performance. Research shows that engaged employees also have a

better health and are absent less often (Demerouti et al., 2001; Schaufeli, Bakker, & Van Rhenen,

2009). Graen, Novak, and Sommerkamp (1982) showed that it is possible to train leaders in their

active listening skills, spending time talking to each subordinate, and sharing expectations.

Compared to the control groups, this training led to gains in LMX quality, job satisfaction, and

productivity. We acknowledge that this may require smaller spans of control and more contact

between leader and subordinates. Besides, is also requires organizations to support their leaders

to invest in their relationship with their followers.

Considering the importance of job resources for improving job performance, it may be

fruitful for organizations to invest in building job resources more formally into the organizational

system. For example, leaders may set up a job enrichment program in which employees are

empowered, while at the same time being supported by their leader, which may provide

followers with opportunities to grow and develop. In addition, leaders may organize a meeting

with each follower at least twice a year, in which followers can talk about the difficulties they

face in their work and discuss with their leader how to solve this. In this way, employees can
 LMX AND JOB PERFORMANCE 15

receive both opportunities for development and social support from the leader. This approach can

also be used when leaders have a large span of control and having a high-quality relationship

with each and every follower is challenging. In this case, all followers benefit from the provision

of resources, because they are more formally built into the organizational system and therefore

available to every employee.

Limitations of the Study

First of all, this is a cross-sectional study, which raises questions about causality. It is also
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

conceivable that employees who are more engaged, have a better relationship with their leader;

likewise, employees who perform better may become more engaged in their work. However, our

results are in line with the motivational process of the JD-R model (Bakker & Demerouti, 2007;

2008), which has also been studied using longitudinal (Hakanen, Perhoniemi, & Toppinen-

Tanner, 2008) and daily diary studies (e.g., Simbula, 2010; Xanthopoulou et al., 2009),

suggesting causality. Addressing the causality issue using a longitudinal design to test the present

study model would nevertheless be a fruitful avenue for future research.

A second limitation is the use of self-reports only, which raises the concern of common

method variance (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003). Although it is unlikely that

common method bias invalidates our findings, because it is rarely strong enough (e.g., Doty &

Glick, 1998; Spector, 2006), we did use Harman’s single-factor test and performed an

exploratory factor analysis. Results show that there is no single factor accounting for the

variance in the data (χ2 = 191.89; df = 10; CFI = .87; RMSEA = .15; SRMR = .10), which makes

the threat of common method bias unlikely. Furthermore, Conway and Lance (2010) argue that

self-reports are appropriate or even the preferred choice in some situations. In the present study,

we were interested in how followers experience their relationship with their leader (i.e., LMX)
 LMX AND JOB PERFORMANCE 16

and how engaged they are, which are private experiences. Next, according to the JD-R model,

each job and each individual has its own constellation of job resources and job demands.

Therefore, followers/employees are the best rater of their job resources. Task performance may

be best measured objectively or by other ratings. However, although far from perfect, self-

reported and leader-rated task performance are moderately related (Bakker & Bal, 2010).

Implications for Future Research and Conclusion

Despite the limitations, this study contributes to the literature by being one of the first to
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

study the mechanism explaining the relationship between LMX and job performance and to

explore the relationship between LMX and employee work engagement. COR theory and the JD-

R model are useful frameworks for continuing this research. For example, having a high-quality

relationship with one’s leader may not only increase job resources, but also valued personal

resources of the employees, like optimism (Tims, Bakker, & Xanthopoulou, 2011), as well as

organization-based self-esteem (OBSE) and meaning-making (Van den Heuvel, Demerouti &

Bakker, 2013). It would be interesting for future research to employ a stronger multi-source

design by examining LMX as reported by the leader or to use leaders’ ratings of employee

performance to prevent common method variance that may influence the results. In a similar

vein, colleague ratings of contextual performance may be used to reduce common method bias

and to examine the process underlying the relationship between LMX and contextual

performance. The same process that was examined in the present study may apply to the

relationship between LMX and contextual performance, especially considering that both LMX

and work engagement have been associated with higher contextual performance (Christian et al.,

2012; Dulebohn et al., 2012).

 LMX AND JOB PERFORMANCE 17

According to the JD-R model, both job resources and job demands are important

predictors of work engagement. In the present study, we only focused on job resources, but

future research may also examine whether having a high-quality relationship with the leader

facilitates challenge demands and prevents hindrance demands. Challenge demands are also

called ‘‘good’’ demands, i.e. demands that promote the personal growth and achievement of the

employee (Podsakoff, LePine, & LePine, 2007), for example workload and time pressure.

Hindrance demands are the ‘‘bad’’ demands that may initiate a health impairment process when
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

they are not compensated with sufficient job resources. Examples are role conflict and role

overload. Research has already shown that LMX is negatively related to hindrance demands

(e.g., Dunegan et al., 2002; Lagace et al., 1993) since leaders in high-quality LMX relationships

take away as many obstacles as possible preventing employees from high performance.

However, there may be a dark side to challenge demands when the quality of the LMX

relationship becomes higher. High-quality LMX relationships are characterized by mutual

obligation, meaning that employees have to return the favors from their leader with exceptional

performance. Eventually, these demands may become overwhelming and act as a source of stress

when workload or time pressures are increasing. Harris and Kacmar (2006) have indeed shown

that the relationship between LMX and stress is best described as curvilinear, whereby

employees in high-quality relationship experience more stress than employees in moderate-

quality relationships. This finding stresses the importance of job resources, since high challenge

demands combined with high job resources are optimal work conditions for employees to thrive,

i.e., being engaged in their work (Tuckey, Bakker, & Dollard, 2012) and thus prevent employees

from experiencing stress. It would be interesting for future research to examine the nature of the
 LMX AND JOB PERFORMANCE 18

relationship between quality of the LMX relationship and challenge demands and the possible

moderating role of job resources in this relationship.

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)
 LMX AND JOB PERFORMANCE 19

References

Bakker, A.B., & Bal, P.M. (2010). Weekly work engagement and performance: A study among

starting teachers. Journal of Occupational and Organizational Psychology, 83, 189-206. doi:

10.1348/096317909X402596.

Bakker, A. B., & Demerouti, E. (2014). Job demands- resources theory. In C. Cooper & P. Chen

(Eds.), Wellbeing: A complete reference guide (pp. 37-64). Chichester, UK: Wiley-

Blackwell.
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Bakker, A.B., & Demerouti, E. (2007). The Job Demands-Resources model: State of the

art. Journal of Managerial Psychology, 22, 309-328. doi: 10.1108/02683940710 73315.

Bakker, A.B., & Demerouti, E. (2008). Towards a model of work engagement. Career

Development International, 13, 209-223. doi: 10.1108/13620430810870476.

Bakker, A. B., Demerouti, E., Taris, T., Schaufeli, W., & Schreurs, P. (2003). A

multigroup analysis of the job demands–resources model in four home care

organizations. International Journal of Stress Management, 10, 16–38. doi:

10.1037/1072-5245.10.1.16.

Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.

Demerouti E., Bakker, A.B., Nachreiner, F., & Schaufeli, W.B. (2001). The Job

Demands-Resources model of burnout. Journal of Applied Psychology, 86, 499-512. doi:

10.1037/0021-9010.86.3.499.

Christian, M.S., Garza, A.S., & Slaughter, J.E. (2011). Work engagement: A quantative review

and test of its relations with task and contextual performance. Personnel Psychology, 64,

89-136. DOI: 10.1111/j.1744-6570.2010.01203.x.

Doty, D.H., & Glick, W.H. (1998). Does common methods variance really bias results?
 LMX AND JOB PERFORMANCE 20

Organizational Research Methods, 1, 374-406. doi: 10.1177/109442819814002.

Dulebohn, J. H., Bommer, W. H. Liden, R. C., Brouer, R. L., & Ferris, G. R. (2012). A

meta-analysis of antecedents and consequences of leader-member exchange: Integrating

the past with an eye toward the future. Journal of Management,38, 1715-1759. doi:

10.1177/0149206311415280.

Dunegan, K.J., Uhl-Bien, M., & Duchon, D. (2002). LMX and subordinate performance: The

moderating effects of task characteristics. Journal of Business and Psychology, 17, 275-
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

286. doi: 10.1177/0149206311415280.

Epitropaki, O., & Martin, R. (2005). From ideal to real: A longitudinal study of the role

of implicit leadership theories on leader-member exchanges and employee outcomes.

Journal of Applied Psychology, 90, 659-676. doi: 10.1037/0021-9010.90.4.659.

Gerstner, C.R., & Day, D.V. (1997). Meta-analytic review of leader-member exchange

theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827-844. doi:

10.1037/0021-9010.82.6.827.

Goodman, S.A., & Svyantek, D.J. (1999). Person-organization fit and contextual

performance: Do shared values matter. Journal of Vocational Behavior, 55, 254-275. doi:

10.1006/jvbe.1998.1682.

Graen, G.B. (1976). Role making process within complex organizations. In M.D.

Dunnette (Ed.), Handbook of industrial organizational psychology. (pp. 1201-1245.)

Chicago: Rand-McNally.

Graen, G., & Cashman, J. (1975). A role-making model of leadership in formal

organization: A development approach. In J. G. Hunt & L. L. Larson (Eds.),

Leadership frontiers. (pp. 143-165). Kent, OH: Kent State University Press.
 LMX AND JOB PERFORMANCE 21

Graen, G.B., Novak, M., & Sommerkamp, P. (1982). The effect of leader-member

exchange and job design on productivity and satisfaction: Testing a dual attachment

model. Organizational Behavior and Human Performance, 30, 109-131. doi:

10.1016/0030-5073%2882%2990236-7.

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, 219-247.

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

doi: 10.1016/1047-9843%2895%2990036-5.

Hakanen, J.J., Perhoniemi, R., & Toppinen-Tanner, S. (2008). Positive gain spirals at

work: From job resources to work engagement, personal initiative and work-unit

innovativeness. Journal of Vocational Behavior, 73, 78-91. doi:

10.1016/j.jvb.2008.01.003..

Halbesleben, J.R.B. (2010). A meta-analysis of work engagement: Relationships with

burnout, demands, resources, and consequences. In M.P. Leiter & A.B. Bakker (Eds.).

Work engagement: A handbook of essential theory and research. (pp. 102-117). New

York: Psychology Press.

Harris, K.J., & Kacmar, K.M. (2006). Too much of a good thing: The curvilinear effect of

leader-member exchange on stress. The Journal of Social Psychology, 146, 65-84. doi:

10.3200/SOCP.146.1.65-84.

Heider, F. (1958). The psychology of interpersonal relationships. Mahwah, NJ, US:

Lawrence Erlbaum Associates Publishers; US.

Hobfoll, S.E. (1989). Conservation of resources: A new attempt at conceptualizing stress.

American Psychologist, 44, 513-524.

 LMX AND JOB PERFORMANCE 22

Hobfoll, S.E. (2002). Social and psychological resources and adaptation. Review of General

Psychology, 6, 307-324. doi: 10.1037//1089-2680.6.4.307.

Hox, J.J. (2010). Multilevel analysis: Techniques and applications (2nd edition). New York, NY,

US: Routledge/Taylor & Francis Group; US.

Hoyle, R.H. (1995). The structural equation modeling approach: Basic concepts and

fundamental issues. In R.H. Hoyle (Ed.), Structural equation modeling: Concepts, issues

and applications (pp. 1-15). Thousand Oaks, CA: Sage.

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Ilies, R., Nahrgang, J.D., & Morgeson, F. (2007). Leader-member exchange and

citizenship behavior: A meta-analysis. Journal of Applied Psychology, 92, 269-277. doi:

10.1037/0021-9010.92.1.269.

Kahn, R.L., Wolfe, D.M., Quinn, R., Snoek, J.D., & Rosenthal, R.A. (1964).

Organizational Stress. New York: Wiley.

Keller, T., & Dansereau, F. (1995). Leadership and empowerment: A social exchange

perspective. Human Relations, 48, 127-146. doi: 10.1177/001872679504800202..

Kline, R.B. (Ed.) (2005). Principles and Practice of Structural Equation Modeling (2nd

ed.). Guilford, New York.

Kop, N., Euwema, M., & Schaufeli, W.B. (1999). Burnout, job stress, and violent

behaviour among Dutch police officers. Work & Stress, 13, 326-340. doi:

10.1080/02678379950019789.

Lagace, R.R., Catleberry, S.B., & Ridnour, R.E. (1993). An exploratory salesforce study of the

relationship between leader-member exchange and motivation, role stress, and manager

evaluation. Journal of Applied Business Research, 9, 110-119.

Le Blanc, P. (1994). De steun van de leiding: Een onderzoek naar het leader member
 LMX AND JOB PERFORMANCE 23

exchange model in de verpleging [Supervisory support: A study on the leader member

exchange model among nurses]. Amsterdam: Thesis Publishers.

Liden, R.C., Sparrowe, R.T., & Wayne, S.J. (1997). Leader-member exchange theory:

The past and potential for the future. In G.R. Ferris (Ed.), Research in Personnel and

Human Resources Management (pp. 47-199). Elsevier Science/JAI Press; US.

Maas, C. J. M., & Hox, J. J. (2005). Sufficient sample sizes for multilevel modeling.

methodology. European Journal of Research Methods for the Behavioral and Social
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Sciences, 1, 85-91. doi: 10.1027/1614-2241.1.3.86.

MacCallum, R.C., Browne, M.W., & Sugawara, H.M. (1996). Power analysis and

determination of sample size for covariance structure modeling. Psychological Methods,

1, 130-149. doi: 10.1037/1082-989X.1.2.130..

Martinussen, M., Richardsen, A.M., & Burke, R.J. (2007). Job demands, job resources

and burnout among police officers. Journal of Criminal Justice, 35, 239-249. doi:

10.1016/j.jcrimjus.2007.03.001.

Muthén, L.K. & Muthén, B.O. (1998-2010). Mplus User’s Guide (Sixth Edition).

Los Angeles, CA: Muthén & Muthén.

Podsakoff, P.M., LePine, J.A., & LePine, M.A. (2007). Differential challenge stressor-

hindrance stressor relationships with job attitudes, turnover intentions, turnover and

withdrawal behavior: A meta-analysis. Journal of Applied Psychology, 92, 438-454. doi:

10.1037/0021-9010.92.2.438..

Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y. & Podsakoff, N.P. (2003). Common method

biases in behavioral research: A critical review of the literature and recommended

 LMX AND JOB PERFORMANCE 24

remedies. Journal of Applied Psychology, 88, 879-903. doi: 10.1037/0021-

9010.88.5.879.

Selig, J. P., & Preacher, K. J. (2008). Monte Carlo method for assessing mediation:

An interactive tool for creating confidence intervals for indirect effects [Computer

software]. Available from http://quantpsy.org/

Scandura, T.A., Graen, G.B., & Novak, M.A. (1986). When managers decide not to

decide autocratically: An investigation of leader-member exchange and decision latitude.

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Journal of Applied Psychology, 71, 579-584. doi: 10.1037/0021-9010.71.4.579.

Scandura T.A., & Williams E.A. (2004). Mentoring and transformational leadership: The role of

supervisory career mentoring. Journal of Vocational Behavior, 65, 448-468. doi:

10.1016/j.jvb.2003.10.003.

Schaufeli, W.B., Bakker, A.B., & Salanova, M. (2006). The measurement of work engagement

with a short questionnaire: A cross-national study. Educational and Psychological

Measurement, 66, 701-716. doi:10.1177/0013164405282471.

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, 893-917. doi: 10.1002/job.595.

Sherony K.M., & Green, S.G. (2002). Coworker exchange: Relationships between

coworkers, leader-member exchange, and work attitudes. Journal of Applied

Psychology, 87, 542-548. doi: 10.1037//0021-9010.87.3.542.

Simbula, S. (2010). Daily fluctuations in teachers’ well-being: a diary study using the Job

Demands-Resources model. Anxiety, Stress, and Coping, 23, 563-584. doi:

10.1080/10615801003728273.
 LMX AND JOB PERFORMANCE 25

Sosik, J. J., & Godshalk, V. M. (2000). Leadership styles, mentoring functions received, and job-

related stress: A conceptual model and preliminary study. Journal of Organizational

Behavior, 21, 365–390. doi: 10.1002/%28SICI%291099-

1379%28200006%2921:4%3C365::AID-JOB14%3E3.0.CO;2-H.

Sparrowe, R. T., & Liden, R. C. (1997). Process and structure in leader-member

exchange. Academy of Management Review, 22, 522-552.

Spector, P. E. (2006). Method variance in organizational research: Truth or urban

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

legend?. Organizational Research Methods, 9, 221-232. doi:

10.1177/1094428105284955.

Tims, M., Bakker, A.B., & Xanthopoulou, D. (2011). Do transformational leaders

enhance their followers’ daily work engagement? The Leadership Quarterly, 22, 121-

131. doi: 10.1016/j.leaqua.2010.12.011.

Townsend, J., Da Silva, N., Mueller, L., Curtin, P., & Tetrick, L. (2002). Attributional

complexity: A link between training, job complexity, decision latitude, leader-member

exchange, and performance. Journal of Applied Social Psychology, 32, 207-221. doi:

10.1111/j.1559-1816.2002.tb01427.x

Tuckey, M.R., Bakker, A.B., Dollard, M.F. (2012). Empowering leaders optimize working

conditions for engagement: A multilevel study. Journal of Occupational Health

Psychology, 17, 15-27. doi: 10.1037/a0025942.

Van den Heuvel, M., Demerouti, E., & Bakker, A.B. (2013). How psychological resources

facilitate adaptation to organizational change. European Journal of Work and

Organizational Psycholog, (ahead-of-print), 1-12. doi: 10.1080/1359432X.2013.817057.

Vasudevan, D. (1993). Developing a model for predicting the career intentions of

 LMX AND JOB PERFORMANCE 26

undergraduate engineering student. Unpublished Ph.D. dissertation, Department of

Management, University of Cincinnati.

Volmer, J., Niessen, C., Spurk, D., Linz, A., & Abele, A.E. (2011). Reciprocal relationships

between leader-member exchange (LMX) and job satisfaction: A cross-lagged analysis.

Applied Psychology: An International Review, 60, 522-545. doi: 10.1111/j.1464-

0597.2011.00446.x

Wrzesniewski, A., & Dutton, J.E. (2001). Crafting a job: Revisioning employees as active
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

crafters of their work. Academy of Management Review, 26, 179-201. doi:

10.2307/259118.

Xanthopoulou, D., Bakker, A.B., Demerouti, E. & Schaufeli, W.B. (2009). Work

engagement and financial returns: A diary study on the role of job and personal

resources. Journal of Organizational and Occupational Psychology, 82, 183-200. doi:

10.1348/096317908X285633.

Xanthopoulou, D., Bakker, A.B., Heuven, E., Demerouti, E., & Schaufeli, W.B. (2008).

Working in the sky: A diary study on work engagement among flight attendants. Journal

of Occupational Health Psychology, 13, 345-356. doi: 10.1037/1076-8998.13.4.345.

Biographical details:

Kimberley Breevaart is a PhD student at the department of Work and Organizational Psychology
at the Erasmus University Rotterdam, The Netherlands. She studied Work and Organizational
Psychology at the Erasmus University. In her PhD project, she focuses on leadership, job
resources, need fulfilment, need for leadership, work engagement and job performance.
Furthermore, she uses dairy designs to examine the daily process through which leaders
influence follower outcomes. Other research interests include burnout and family-work
interference. She is a member of the European Association of Work & Organizational
Psychologists (EAWOP) and has reviewed for various scientific journals including European
 LMX AND JOB PERFORMANCE 27

Journal of Work and Organizational Psychology, Applied Psychology: An International Review,

and Anxiety, Stress, and Coping.

Arnold B. Bakker is professor and chair of the department of Work & Organizational
Psychology, Erasmus University Rotterdam, The Netherlands, and Adjunct professor at the
department of Sociology and Social Policy, Lingnan University, Hong Kong. He is also president
of the European Association of Work and Organizational Psychology, and fellow of the
Association for Psychological Science. Bakker’s research interests include positive
organizational phenomena such as work engagement, flow, and happiness at work. See also:
www.arnoldbakker.com.

Evangelia Demerouti is a full professor of organizational behavior at Eindhoven University of

Technology, The Netherlands. She studied psychology at the University of Crete and received
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

her PhD in the Job Demands-Resources model of burnout (1999) from the Carl von Ossietzky
Universität Oldenburg, Germany. Her main research interests concern topics from the field of
work and health including the Job Demands – Resources model, burnout, work-family interface,
crossover of strain, flexible working times, and job performance. She has published over 100
national and international papers and book chapters on these topics, is associate editor of the
Journal of Personnel Psychology, serves as a reviewer for various national and international
scientific journals. Her articles have been published in journals including Journal of Vocational
Behavior, Journal of Occupational Health Psychology, and Journal of Applied Psychology.

Machteld van den Heuvel (Maggie) is an Assistant Professor of Work and Organizational
Psychology at University of Amsterdam, The Netherlands. She studied Occupational Health
Psychology at Utrecht University. After some years working as an organizational psychologist in
practice, she went back and completed her PhD at Utrecht University. Her research focuses on
personal resources, meaning-making, job crafting, work engagement and adaptation to change.
Interventions studies with the aim of improving employee well-being at work are another area of
interest. Maggie combines her academic work with working in the field as a trainer / coach (see:
www.artofwork.nl). She holds a BPS Certificate of Competence in Occupational Testing. She is
a member of the European Association of Work & Organisational Psychologists (EAWOP) and
serves as a reviewer for various national and international scientific journals. E-mail:
m.vandenheuvel2@uva.nl
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Autonomy

LMX1 A1 A2 A3 A4

LMX2
Developmental Work Jo
LMX3 LMX
Opportunities Engagement Perform
LMX4

LMX5
D1 D2 D3 V1-V3 A1-A3 D1-D3

Social
Support

S1 S2 S3 S4

Figure 1. Graphical representation of the hypothesized model.

 .15***

Autonomy

.40*** .07
LMX1 A1 A2 A3 A4

LMX2 J1
.38*** .25*** .34***
Developmental Work Job
LMX3 LMX J2
Opportunities Engagement Performance
LMX4 J3

LMX5
D1 D2 D3 V1-V3 A1-A3 D1-D3
Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

.51*** .37***

Social
Support

S1 S2 S3 S4

Figure 2. The process model of leader-member exchange showing standardized

estimates.
 Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

Table 1

Means, Standard Deviations, Inter-Correlations and Internal Consistencies (Conbrach’s alphas on the diagonal) Between the Study Variables,

N = 58 Teams, N = 847 Employees).

M SD 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

1. Gender 1.36 .48 -

2. Age 42.68 10.34 -.29*** -

3. Education 3.16 1.06 .04 -.05 -

4. Working hours 35.38 7.31 -.43*** -.07* .10*** -

5. LMX 3.03 .89 .07* .05 .03 -.01 (.92)

6. Developmental Opportunities 3.46 .81 .09* .08* .04 .14*** .51*** (.87)

7. Social Support 3.88 .77 .003 -.20*** -.04 -.02 .38*** .24*** (.89)

8. Autonomy 3.27 .76 .02 .17*** .20*** .11*** .40*** .51*** .19*** (.81)

9. Work Engagement 3.95 .95 .06 .02 -.06 .13*** .46*** .55*** .41*** .36*** (.95)

10. Job Performance 5.02 .50 .10* .11*** .05 .12** .17*** .19*** .10** .20*** .34*** (.86)
 Table 2.

Fit of the Measurement Models.

Fit indices

Models CFI RMSEA SRMR

1. One-factor model .51 .16 .14

2. Four-factor model .78 .11 .09

3. Five factor model .84 .09 .08

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

4. Six-factor model .93 .06 .04

Note. In model 2 all job resources were combined into a single factor. In model three
LMX and social support were combined into a single factor.
 Table 3.

Maximum Likelihood Estimates, Standard Errors, and Confidence Intervals for the

Direct, Indirect and Contrast Effects (N = 58 Teams, N = 847 Employees).

Unstandardized 95% CI

Indirect Effects Est. S.E. p Lower Upper

1. LMX
Autonomy
WE .03 .02 n.s. -.001 .06

2. LMX
Developmental Opportunities
WE .20 .03 p < .001 .16 .25

Downloaded by UNIVERSITY OF EXETER At 06:59 08 August 2015 (PT)

3. LMX
Social Support
WE .10 .02 p < .001 .07 .14

Contrast Effects

Indirect Effect 1 – Indirect Effect 2 -.18 .03 p < .001 -.25 -.10

Indirect Effect 1 – Indirect Effect 3 -.07 .02 p < .05 -.12 -.03

Indirect Effect 2 – Indirect Effect 3 -.10 .03 p < .001 -.16 -.04

Indirect Effects

1. LMX
Developmental Opportunities
WE

.04 .01 p < .001 .02 .05

Performance

2. LMX
Social Support
WE
Performance .02 .00 p < .001 .01 .03

Contrast Effect

Indirect Effect 1 – Indirect Effect 2 .02 .01 p < .001 .01 .03