Journal of Nursing Management, 2011, 19, 41–49

The mediating role of job satisfaction between leader-member

exchange and turnover intentions

1 2
GUOHONG (HELEN) HAN PhD and MARC JEKEL Dipl.-Psych.
1
Assistant Professor, Department of Management, Williamson College of Business Adminstration, Youngstown State
University, Youngstown, OH, USA and 2Research Fellow, Department of Psychology, University of Bonn, Bonn,
Germany

Correspondence H A N G . ( H E L E N ) & J E K E L M . (2011) Journal of Nursing Management 19, 41–49

Guohong (Helen) Han The mediating role of job satisfaction between leader-member exchange and
Department of Management turnover intentions
Williamson College of Business
Adm. Aim The aim of the present study was to investigate whether job satisfaction
Youngstown State University mediates between leader-member exchange and nurse turnover intentions.
One University Plaza Background Limited knowledge is available on the mediating role of job satisfac-
Youngstown, OH 44555 tion between leader-member exchange and turnover intentions in the nursing field.
USA Method This is a cross-sectional survey study. Data were collected in a US hospital.
E-mail: ghan@[Link] Results Higher leader-member exchange was associated with lower turnover inten-
tions (b = )0.41, P < 0.001). Leader-member exchange was positively related to oneÕs
job satisfaction (b = 0.50, P < 0.001). Job satisfaction was negatively related to
turnover intentions (b = )0.64, P < 0.001). When leader-member exchange and job
satisfaction were both controlled for, the link between leader-member exchange and
turnover intentions was no longer significant (b = )0.12, P > 0.05).
Conclusion Job satisfaction mediates the link between leader-member exchange and
turnover intentions.
Implications for nursing management The present study reinforces the significant
role of managers. Nurse managers can try to cultivate a good relationship with the
nurses. This can enhance nursesÕ job satisfaction and in turn lower their turnover
intentions.
Keywords: job satisfaction, LMX, turnover intentions

Accepted for publication: 23 August 2010

difficult and expensive to replace (e.g. Waldman et al.

Introduction
2004). A variety of variables have been found to predict
An abundance of studies (e.g. McGowan 2001, Trink- nurse turnover. These include poor or inflexible work
off et al. 2006) have been devoted to understanding schedules (Simons 2008), work stress (Ramrup & Pacis
nursing shortages in terms of what contributes to nurse 2008) and heavy workloads (McGowan 2001). Turn-
turnover, a very costly problem that will continue along over intentions, meaning the conscious and deliberate
with the substantial global nurse shortage (World willingness to leave the organization (Tett & Meyer
Health Organization 2006). Retention is critical for the 1993), have been shown to be highly correlated with
health-care sector, where employees who leave are actual turnover (Hom & Hulin 1981).
DOI: 10.1111/j.1365-2834.2010.01184.x
ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd 41
G. Han and M. Jekel

Researchers have argued that there has been a paucity accordingly (Thibault & Kelley 1959). High-quality
of research on turnover intentions despite its critical LMX employees are more likely to get attention and
importance in predicting actual turnover (McCarthy support from their supervisors as an exchange for their
et al. 2007). In order to address this gap, the present hard work. This kind of social exchange will ultimately
study focuses on turnover intentions and particularly lead to higher job satisfaction (Podsakoff et al. 1996).
how leadership and job satisfaction can work syner- Empirical studies have confirmed that high-quality
gistically to impact turnover intentions in the field of LMX contributes to favourable employee attitudes and
nursing. behaviours, including organizational citizenship behav-
Employees usually do not quit their companies, they iours (e.g. Deluga 1994), higher performance (Dunegan
quit their bosses instead (Smith 2000). Leader-member et al. 2002) and organizational commitment (Scandura
exchange (LMX) refers to the quality of the relationship & Graen 1984, Dienesch & Liden 1986). Research has
between an individual and his/her supervisor (Graen also demonstrated that LMX is positively related to
1976). In spite of the number of research studies employee job satisfaction (Graen et al. 1982a,b, Scan-
exploring the consequences of LMX, the link between dura & Graen 1984, Sparrow 1994, Gerstner & Day
LMX and turnover intentions has been unclear and 1997).
ÔequivocalÕ (Gerstner & Day 1997). Gerstner and Day In the field of nursing, Tangirala et al. (2007) found
(1997) called for research to examine more closely by that high-quality LMX was positively related to sub-
looking at variables that mediate or moderate this ordinatesÕ work attitudes. Spence-Laschinger et al.
process and proposed that LMX might affect turnover (2001) also found that subordinates would display po-
through work attitudes, including satisfaction (Gerstner sitive work attitudes if supervisors provided them with a
& Day 1997). supportive work environment. Another study by Chen
Responding to their call, the present paper aims to et al. (2008) found that LMX was significantly associ-
study the mediating role of job satisfaction on the ated with nursesÕ perception of trust in their supervisors
relationship of LMX and turnover intentions. To the and ultimately their engagement in organizational citi-
best of our knowledge, the present study will be the first zenship behaviours. Likewise, LaschingerÕs (2007) study
study that investigates the above mediation model in the on 141 hospital-based nurse managers found the sig-
nursing field. We intend to address the following re- nificant and positive correlation between LMX and job
search questions: Will high-quality LMX lead to higher satisfaction.
job satisfaction? If so, will this lead to lower turnover Hence, our first hypothesis is:
intentions?
H1: LMX is positively related to oneÕs job satis-
faction.
Conceptual framework and hypotheses
The link between LMX and job satisfaction The link between job satisfaction and turnover
intentions
LMX researchers argue that leaders display different
leadership behaviours when dealing with different Drawing upon the theory of planned behaviour (Ajzen
subordinates (e.g. Graen & Cashman 1975). High- 1985, 1991), oneÕs attitudes towards a specific behav-
quality LMX individuals go the extra mile and con- iour, subjective norms and perceived behavioural con-
tribute more towards work success. As a result, they trol are the three building blocks of a behavioural
receive greater supervisor attention and more support. intention that initiates actual behaviour. Employees
Low-quality LMX individuals, on the other hand, do who are not satisfied with their jobs will experience
not enjoy these preferential treatments and have a more negative attitudes towards their jobs and positive atti-
contractual relationship with the supervisor (e.g. Graen tudes towards the intention to quit the job. If employees
& Cashman 1975). think they have resources to control the behaviour (e.g.
Locke (1976) defined job satisfaction as Ôa pleasurable skill to quit and find another job) and if the turnover
or positive emotional state resulting from the appraisal behaviour does not conflict with subjective norms (e.g.
of oneÕs job or job experiencesÕ (p. 1304). Social ex- behaviour that they believe to lead to social punishment
change theory (Blau 1964) holds that two or more by significant others), they will likely foster turnover
parties exchange with each other intangible social costs intentions that will finally motivate turnover behaviour.
and benefits, such as respect, honour, friendship and In research, job satisfaction has been found to be a
caring, assuming that the other party will collaborate significant predictor of turnover intentions (Van Dick

42 ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49
 LMX, job satisfaction and turnover intentions

et al. 2004, Wright & Bonett 2007, Lee et al. 2008). turnover intentions as well as actual turnover. High-
Additionally, Cotton and Tuttle (1986) found that quality LMX subordinates are, therefore, more likely to
overall job satisfaction was negatively related to turn- be satisfied with their jobs overall and less likely to
over. think about quitting.
Similar findings have also been established in the The negative association between LMX and turnover
nursing field. Price and Mueller (1981) proposed and intentions has also solicited support from nursing field
empirically tested the causal model of turnover, in which research. Laschinger et al. (2009) found that LMX can
job satisfaction affects intention to leave and ultimately predict nursesÕ organizational commitment. Likewise,
turnover. On the basis of longitudinal data on 1091 Brunetto et al. (2010) found that the supervisor–sub-
registered nurses (RNs) in seven hospitals, they found ordinate relationship was positively associated with
that RNs with higher job satisfaction are more likely to nursesÕ affective commitment. Obviously, this would
stay and less likely to have intentions to leave (Price & indicate low intentions of turnover. Another study (Le
Mueller 1981). In a survey of nurses from 12 units in a Blanc et al. 1993) found that LMX was negatively re-
908-bed university hospital in the Southwest, Shader lated to oneÕs propensity to quit the job.
et al. (2001) observed the significant negative relation- Therefore, it seems reasonable to predict the negative
ship between job satisfaction and the anticipated turn- relationship between LMX and oneÕs turnover inten-
over. Additionally, McCarthy et al. (2007) conducted a tions. As we have hypothesized the positive relationship
survey on 352 RNs at 10 hospital sites throughout the between LMX and job satisfaction in H1 and the neg-
Republic of Ireland and found the significant association ative relationship between job satisfaction and turnover
between job satisfaction and intention to stay. Based on intentions in H2, we can now predict that:
the above, our second hypothesis is as follows:
H3: Job satisfaction mediates between LMX and
H2: Job satisfaction is negatively related to turn- turnover intentions.
over intentions.

Methods
The mediating effect of job satisfaction between
The studied organization is a region of Catholic
LMX and turnover intentions
Healthcare Partners, which is the largest health system
High LMX employees tend to have a trust-based rela- in one mid-western state in the United States. This non-
tionship with their supervisors (Graen & Uhl-Bien profit health care provider offers a range of services,
1995), and consequently they get more resources and including skilled nursing and independent living.
information-based support from their supervisors. Thanks to the continuous effort of the management, the
High-quality LMX individuals belong to in-group studied organization saw a steady decrease in nurse
members and accordingly enjoy preferential support, turnover over the past few years. However, since 2007,
for instance, being assimilated into the leaderÕs personal the organization had been experiencing another round
network (Sparrow & Liden 2005). So high-quality of high turnover. The present study was conducted
LMX individuals are inclined to feel more accepted and during the fall of 2008. Data collection lasted for about
valued. Hence, they are more likely to reap individual 2 months. This paper used a subset of data from the
and organizational benefits. Many studies have proved project which aims to investigate what factors can help
significant and positive consequences of having a high- to reduce nurse turnover rates.
quality LMX. To illustrate, Sagas and Cunningham The first author of the study sent out a letter, a re-
(2004) found that high LMX people reported greater search proposal and an Institutional Review Board
job and career satisfaction. Wayne et al. (1997) found permission letter to the Vice President of the Human
that LMX is positively related to perceived organiza- Resources Department. The letter and the research
tional support, which contributes to employee satis- proposal presented the main aims of the study to
faction (Matzler & Renzl 2006) and satisfying examine the ways to assess leadership and employee
employeesÕ socioemotional needs (Cohen & Prusak work attitudes in a diverse working environment. The
2001). LMX has also been found to enhance work letter also assured that the study would follow strict
motivation (Lagace et al. 1993) and employeesÕ well- guidelines of confidentiality.
being at work (Sparr & Sonnentag 2008). In a longi- Based on our request, the Human Resources Depart-
tudinal study of 116 new executives, Bauer et al. (2006) ment selected seven hospital units. Two of them have an
found that LMX is negatively related to new executive excellent retention rate with very little attrition whereas

ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49 43
G. Han and M. Jekel

the other five have witnessed an increase in turnover with their organization. CronbachÕs alpha was 0.92
rates. The coordinator of the Human Resources De- (Chen et al. 2004). Respondents were asked to use a
partment communicated with every nurse manager of five-point Likert scale ranging from 1 = strongly dis-
the seven units. The surveys were then collected on site agree to 5 = strongly agree. Items included ÔIÕm satisfied
when the nurses were changing their shifts. with the degree of challenge that my job providesÕ; ÔIÕm
satisfied with the chances/opportunities for promotionÕ;
ÔIÕm satisfied with the relationship I have with my
Measures supervisorÕ; ÔIÕm satisfied with the relationship I have
This study controlled for the following demographic with my co-workersÕ; ÔIÕm satisfied with the level of pay
differences: age, gender, education and organization that my job providesÕ; ÔIÕm satisfied with the amount of
tenure. recognition I get for good work (my job, overall)Õ. A
summed average of the six items was produced to form
Turnover intentions the job satisfaction score.
We used McKay et al.Õs (2007) scale to measure turnover
intentions. Sample items included ÔI hardly ever think
about leavingÕ; and ÔIt would take a lot to get me to leave Analysis
the companyÕ. Responses to the scale were scored n a six- We analysed data using the software package R (R De-
point Likert scale ranging from 1 = strongly disagree to velopment Core Team, 2010). R is a free statistical
6 = strongly agree. Intention scores were recoded so that analysis software and can be downloaded from the
low scores denote lower intention to leave the firm. ÔComprehensive R Archive NetworkÕ at [Link]
CronbachÕs alpha was 0.90 (McKay et al. 2007). [Link]/. To learn more about R , we propose the
free ebook introduction to R by Paradis (2005).
LMX We evaluated descriptive statistics to describe char-
We used Graen et al. (1982a,b) seven-item LMX mea- acteristics of our participants. We excluded participants
sure. Studies have shown that it has sound psychometric with missing data on any of the measures (n = 16). We
validity (e.g. Graen & Uhl-Bien 1995, Gerstner & Day evaluated CronbachÕs alpha as reliability estimates
1997). This variable was assessed using seven items across all subscales. Statistical analyses included testing
based on the member version of LMX questionnaires the correlation (PearsonÕs r) among the various vari-
developed and used before the research (e.g. Liden & ables and running ordinary multivariate linear regres-
Graen 1980, Wayne et al. 1997). CronbachÕs alpha was sions. The mediation model was tested using the most
0.93 (Graen et al. 1982a,b). Respondents were asked to prevalent method (Baron & Kenny 1986) in testing
indicate the extent to which the items characterize the mediation. To illustrate, Baron and Kenny (1986) ar-
quality of their exchange relationships with their gued that a mediation model needs to satisfy the fol-
supervisors, using a seven-point Likert scale ranging lowing four steps: (1) there is a significant effect
from 1 = strongly disagree to 7 = strongly agree. Items between the independent variable and the mediator; (2)
included ÔMy supervisor would be personally inclined to there is a significant effect between the mediator and the
help me solve problems in my workÕ; ÔMy working dependent variable; (3) there is a significant effect be-
relationship with my supervisor is effectiveÕ; ÔI have tween the independent variable and the dependent
enough confidence in my supervisor that I would defend variable; and (4) the significant relationship between the
and justify his/her decisions if he/she was not present to independent and the dependent variable is reduced
do soÕ; ÔMy supervisor considers my suggestions for (=partial mediation) or removed (=full mediation) after
changeÕ; ÔMy supervisor and I are suited to each otherÕ; controlling for the mediator. In addition to this, we also
ÔMy supervisor understands my problems and needsÕ; conducted the Sobel test based on Preacher and Leo-
ÔMy supervisor recognizes my potentialÕ. The answers nardelliÕs (2001) procedure to confirm the significance
were averaged to get the final score for LMX. of the mediation model.

Job satisfaction
According to Chen et al. (2004), we asked the partici- Results
pants to indicate their level of satisfaction with the
Participants
following aspects of their present job: (1) job content,
(2) supervision, (3) coworker relations, (4) opportuni- The survey was distributed to about 400 nurses from
ties for promotion, (5) pay and (6) overall satisfaction seven units of the hospital in three locations. One

44 ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49
 LMX, job satisfaction and turnover intentions

hundred and ninety-seven nurses participated in the Impact of job satisfaction on turnover intentions
present study. The response rate was 49.2%. We re-
Job satisfaction was found to be negatively related to
moved 16 individuals owing to missing data. Thus, the
turnover intentions (b = )0.64, P < 0.001, Table 3).
final sample size was 181. ParticipantsÕ demographics
Therefore, it supported Hypothesis 2. It appeared that
are displayed in Table 1. Most participants were female
higher job satisfaction results in lower turnover inten-
(91% of all participants), more than 35 years old (62%
tions.
of all participants) and held a high school degree (42%
of all participants).
Mediating effect of job satisfaction between LMX
and turnover intentions
Descriptive statistics
The mediation was tested by adopting the procedure
In the present study, CronbachÕs alphas for LMX
suggested by Baron and Kenny (1986). As displayed in
(mean = 5.15, SD = 1.56), job satisfaction (mean = 3.5,
Table 4, regression results showed that LMX was neg-
SD = 0.87) and turnover intentions (mean = 3.45,
atively linked to turnover intentions (b = )0.41, P <
SD = 1.47) were 0.97, 0.84 and 0.90, respectively.
0.001). When job satisfaction was added in (see
Participants higher in LMX were more satisfied with
Table 5), LMX became insignificant (b = 0.12, P >
their jobs (PearsonÕs r = 0.5, P < 0.01) and showed less
0.05). However, job satisfaction was significantly and
turnover intentions (PearsonÕs r = )0.4, P < 0.01).
negatively related to turnover intentions (b = )0.58,
Furthermore, participants with higher job satisfaction
P < 0.001). This fully supported Hypothesis 3 about the
showed less turnover intentions (PearsonÕs r = )0.59,
mediating effect of job satisfaction between LMX and
P < 0.01).
turnover intentions.
To further test the mediation effect, the Sobel test was
Impact of LMX on job satisfaction conducted according to Preacher and LeonardelliÕs
(2001) method. The results indicated that the indirect
Based on linear regression, LMX was found to have a
significant and positive effect on job satisfaction Table 2
(b = 0.50, P < 0.001, Table 2). Thus, Hypothesis 1 Regression results of leader-member exchange (LMX) on job satis-
faction
received support. It appeared that higher-quality LMX
results in higher job satisfaction. Variables Standardized betas Standard error

Intercept )0.56 0.31

Focus variable
LMX 0.50*** 0.06
Table 1
Control variables
ParticipantsÕ demographics
Gender
Categorial variables n % Female 0.28 0.22
Age group
Gender 26–30 years 0.07 0.27
Female 164 91 31–35 years )0.29 0.25
Male 17 9 36–40 years 0.31 0.29
Age group (in years)* 41–45 years 0.11 0.25
20–25 19 10 46–50 years 0.14 0.26
26–30 22 12 51–55 years 0.31 0.30
31–35 27 15 56–60 years 0.74* 0.33
36–40 16 9 >60 years 0.93 0.64
41–45 35 19 Education
46–50 29 16 Associate degree 0.35* 0.16
51–55 20 11 BachelorÕs degree 0.36* 0.15
56–60 11 6 MasterÕs degree 0.01 0.51
>60 2 1 Organization tenure )0.06 0.07
Education Summary statistics
High school 76 42 R2 0.35
Associate degree 45 25 Adjusted R2 0.29
BachelorÕs degree 57 31 F14,166 6.26***
MasterÕs degree 3 2
Continuous variable Mean SD *P < 0.05, ***P < 0.001; Gender, age group and education were
Organization tenure (in years) 6.09 7.19 dummy coded, i.e., the intercept is the mean of the standardized job
satisfaction for male participants of the age group 20–25 years with a
*Percentages do not add to 100% as a result of rounding. high school degree.

ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49 45
G. Han and M. Jekel

Table 3 Table 4
Regression results of job satisfaction on turnover intentions Regression results of leader-member exchange (LMX) on turnover
intentions
Variables Standardized betas Standard error
Variables Standardized betas Standard error
Intercept )0.87** 0.28
Independent variables Intercept )0.53 0.33
Job satisfaction )0.64*** 0.07 Focus variable
Control variables LMX )0.41*** 0.07
Gender Control variables
Female 0.49* 0.20 Gender
Age group Female 0.32 0.24
26–30 years 0.07 0.24 Age group
31–35 years 0.40 0.23 26–30 years 0.12 0.29
36–40 years 0.29 0.27 31–35 years 0.61* 0.27
41–45 years 0.27 0.23 36–40 years 0.10 0.31
46–50 years 0.68** 0.24 41–45 years 0.19 0.27
51–55 years 0.09 0.27 46–50 years 0.61* 0.28
56–60 years 0.78* 0.30 51–55 years )0.05 0.32
>60 years 0.82 0.59 56–60 years 0.30 0.35
Education >60 years 0.28 0.68
Associate degree )0.03 0.15 Education
BachelorÕs degree 0.34* 0.14 Associate degree )0.26 0.17
MasterÕs degree )0.14 0.46 BachelorÕs degree 0.11 0.17
Organization tenure )0.10 0.07 MasterÕs degree )0.07 0.55
Summary statistics Organization tenure )0.07 0.08
R2 0.46 Summary statistics
Adjusted R2 0.41 R2 0.25
F14,166 9.90*** Adjusted R2 0.18
F14,166 3.88***
*P < 0.05, **P < 0.01, ***P < 0.001; Gender, age group and edu-
cation were dummy coded, i.e., the intercept is the mean of the *P < 0.05, ***P < 0.001; Gender, age group and education were
standardized turnover intentions for male participants of the age dummy coded, i.e., the intercept is the mean of the standardized
group 20–25 years with a high school degree. turnover intentions for male participants of the age group 20–
25 years with a high school degree.
effect as a result of mediation was significant
(Z = )5.64, P < 0.001). This confirmed the full medi- related to turnover intentions via the mediator of job
ation effect of job satisfaction between LMX and satisfaction. The positive association between LMX and
turnover intentions. job satisfaction is consistent with the current research
For an overview of the central results reported, the (e.g. Sparrow 1994, Laschinger 2007). If employees
mediated and unmediated model with the respective b have a good quality relationship with their bosses, they
weights is displayed in Figure 1. In the upper part of the are more likely to be happy with their jobs. The nega-
figure, b weights for the mediated model are displayed, tive linkage between job satisfaction and turnover
i.e. b weights for the link between LMX and job satis- intentions provides further support to the extant liter-
faction (b = 0.50), job satisfaction and turnover inten- ature (e.g. Shader et al. 2001, McCarthy et al. 2007;). If
tions (b = )0.64) and LMX and turnover intentions employees are happy with their work overall, they are
when controlled for job satisfaction (b = )0.12) are less likely to contemplate quitting. As previously ar-
displayed. In the lower part of the figure, the b weight gued, to the best of our knowledge, this is the first study
for the unmediated model is displayed, i.e. the b weight that has examined the mediating role of job satisfaction
for the link between LMX and turnover intentions between LMX and turnover intentions in the nursing
(b = )0.41). In total, the model with all variables in- field. Therefore, cultivating and enhancing nursesÕ job
cluded explained 42% of the variance in turnover satisfaction plays a crucial rule in nurse retention. These
intentions and the mediator (job satisfaction) explained findings have both theoretical and practical implica-
an additional 24% unique variance. tions.

Discussion Theoretical implications

We made a contribution to the LMX literature by The present study contributes to the leadership
introducing a potential mediator between LMX and literature. Theoretically, LMX has been shown to be
turnover intentions. We found that LMX is negatively related to a number of favourable organizational and

46 ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49
 LMX, job satisfaction and turnover intentions

Table 5
Regression results of leader-member exchange (LMX) on turnover
intentions (controlling for job satisfaction)

Variables Standardized betas Standard error

Intercept )0.85** 0.28

Focus variables
LMX )0.12 0.07
Job satisfaction )0.58*** 0.07
Control variables
Gender
Female 0.48* 0.20
Age group
26–30 years 0.09 0.24
31–35 years 0.44 0.23
36–40 years 0.28 0.27
41–45 years 0.26 0.23
46–50 years 0.70** 0.24
51–55 years 0.13 0.27
56–60 years 0.73* 0.30
>60 years 0.82 0.58
Education
Associate degree )0.06 0.14 Figure 1
BachelorÕs degree 0.32* 0.14 Mediating role of job satisfaction between leader-member exchange
MasterÕs degree )0.06 0.46 and turnover intentions. Displayed is the mediated (upper part) and
Organization tenure )0.10 0.06 unmediated (lower part) linear model.
Summary statistics
R2 0.47 leader receives feedback from multi-sources, which is
Adjusted R2 0.42 named Ô360 degree feedbackÕ (Atwater et al. 2007), can
F14,166 9.57***
be utilized to improve nurse managersÕ leadership skills
*P < 0.05, **P < 0.01, ***P < 0.001; Gender, age group and edu- (Grant et al. 2009, Karsten 2010). For example, nurse
cation were dummy coded, i.e., the intercept is the mean of the managers can be assessed in terms of LMX via the rat-
standardized turnover intentions for male participants of the age ings of their subordinates to diagnose their leadership
group 20–25 years with a high school degree.
behaviour. On the basis of the results, nurse managers
can modify their behaviour accordingly.
individual outcomes. However, there has not been a These findings also help us to better understand how
clear picture about the impact of LMX on turnover. to reduce turnover rates among nurses. Nurses and
Responding to the researchersÕ call (Gerstner & Day nurse aides make up a large percentage of the workers
1997) for the potential mediator between LMX and in the healthcare system in the United States. However,
turnover, this study found that having a quality rela- there have been a lot of concerns about the shortage of
tionship with oneÕs boss does help the organization to nurses over the last couple of years in spite of the overall
retain talent by enhancing employeesÕ job satisfaction. unfavourable economy. The United States is expected to
have a shortage of about 340 000 nurses by the year
2020 (Auerbach et al. 2007). Research showed one in
Practical implications for nursing management
five nurses reporting an intention to leave their job
Practically, the study reinforced the significance of the within a year (Aiken et al. 2001, 2002). If those inten-
role of managers. This study found that cultivating a tions are carried out, it can be an alarming prospect for
good relationship with staff nurses can greatly help to the health care work force in America. The shortage
reduce oneÕs intention to leave. Just as Andrews and will not only make the hospitals suffer as a whole, but
Dziegielewski (2005) pointed out, given the time con- also cause a great deal of inconvenience to patients. A
straint, that the first-line nurse managers, should make nursing shortage coupled with the aging population
an attempt to identify factors that precipitate low satis- with an increasing need for healthcare creates a Ôperfect
faction. This undoubtedly is very challenging to nurse stormÕ (Hernandez 2005), which will result in an
managers, especially against the background of the di- unprecedented difficult situation. The present study
verse workforce (Carver & Candela 2008). Diversity makes an effort to investigate how to retain nurses by
training should be provided to nurse managers in terms highlighting the importance of having a good relation-
of how to cater to different peopleÕs needs and raise their ship with their bosses and reinforces the role of job
level of job satisfaction. The rating method in which the satisfaction in oneÕs turnover intentions. Management

ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49 47
G. Han and M. Jekel

hence needs to be trained to work on cultivating good Brunetto Y., Farr-Wharton R. & Shacklock K. (2010) The impact
relationships with employees and increasing employeesÕ of supervisor-subordinate relationships on moral: implications
for public and private sector nursesÕ commitment. Human Re-
perception of job satisfaction.
source Management Journal 20, 206–225.
Carver L. & Candela L. (2008) Attaining organizational com-
Limitations mitment across different generations of nurses. Journal of
Nursing Management 16, 984–991.
Notwithstanding, this study has a few limitations that Chen T.Y., Chang P.L. & Yeh C.W. (2004) An investigation of
should be noted. First, this study was a cross-sectional career development programs, job satisfaction, professional
development and productivity: the case of Taiwan. Human
study, which did not provide us with confidence about
Resource Development International 7, 441–463.
the causality of the model. Longitudinal research is thus Chen C.-H.V., Wang S.-J., Chang W.-C. & Hu C.-S. (2008) The
called for in the future to investigate this. Second, the effect of leader-member exchange, trust, supervisor support on
study was based on self-report survey responses, which organizational citizenship behavior in nurses. The Journal of
was liable to common method variance bias. Nursing Research 16, 321–328.
To conclude, we found that high-quality LMX can Cohen D. & Prusak L. (2001) In Good Company: How Social
Capital Makes Organizations Work. Harvard Business School
lead to greater job satisfaction, which in turn results in
Press, Boston, MA.
less turnover intentions. We made a great effort to Cotton J. & Tuttle J. (1986) Employee turnover: a meta-analysis
contribute to the nursing research by examining the and review with implications for research. Academy of Man-
mediating role of job satisfaction between LMX and agement Review 11, 55–70.
turnover intentions. Given the poor economy and bud- Deluga R.J. (1994) Supervisor trust building, leader-member ex-
change and organizational citizenship behavior. Journal of
get constraint, fostering a good relationship between a
Occupational and Organizational Psychology 67, 315–326.
nurse manager and staff nurses might be a cost-effective Dienesch R.M. & Liden R.C. (1986) Leader-member exchange
and efficient strategy to cut back on nurse turnover rate. model of leadership: a critique and further development.
Academy of Management Review 11, 618–634.
Dunegan K.J., Uhl-Bien M. & Duchon D. (2002) LMX and
References subordinate performance: the moderating effects of task char-
Aiken L.H., Clarke S.P., Sloane D.M. et al. (2001) NurseÕ reports acteristics. Journal of Business and Psychology 17, 275–285.
on hospital care in five countries. Health Affairs 20, 43–53. Gerstner C.R. & Day D.V. (1997) Meta-analytic review of leader-
Aiken L.H., Clarke S.P., Sloane D.M., Sochalski J. & Silber J.H. member exchange theory: correlates and construct issues.
(2002) Hospital nurse staffing and patient mortality, nurse Journal of Applied Psychology 82, 827–844.
burnout, and job dissatisfaction. Journal of the American Graen G.B. (1976) Role making processes within complex orga-
Medical Association 288, 1987–1993. nizations. In Handbook of Industrial and Organizational Psy-
Ajzen I. (1985) From intentions to actions: a theory of planned chology (M.D. Dunnette ed.), pp. 1201–1245. Rand McNally,
behavior. In Action Control: From Cognition to Behavior Chicago, IL.
(J. Kuhl & J. Beckmann eds), pp. 11–39. Springer-Verlag, Graen G.B. & Cashman J.F. (1975) A role making model in
Heidelberg. formal organizations: a developmental approach. In Leadership
Ajzen I. (1991) The theory of planned behavior. Organizational Frontiers (J.G. Hunt & L.L. Larson eds), pp. 143–165. Kent
Behavior and Human Decision Processes 50, 179–211. State Press, Kent, OH.
Andrews D.R. & Dziegielewski S.F. (2005) The nurse manager: Graen G.B. & Uhl-Bien M. (1995) Relationship-based approach
job satisfaction, the nursing shortage and retention. Journal of to leadership: development of leader-member exchange (LMX)
Nursing Management 13, 286–295. theory of leadership over 25 years: applying a multilevel
Atwater L.E., Brett J.F. & Charles A.C. (2007) Multisource multidomain perspective. Leadership Quarterly 6, 219–247.
feedback: lessons learned and implications for practice. Human Graen G.B., Liden R.C. & Hoel W. (1982a) Role of leadership in
Resource Management 46, 285–307. the employee withdrawal process. Journal of Applied Psy-
Auerbach D., Buerhaus P. & Staiger D. (2007) Better late than chology 67, 868–872.
never: workforce supply implications of late entry into nursing. Graen G.B., Novak M.A. & Sommerkamp P. (1982b) The effects
Health Affairs 26, 178–185. of leader-member exchange and job design on productivity and
Baron R.M. & Kenny D.A. (1986) The moderator-mediator satisfaction: testing a dual attachment model. Organizational
variable distinction in social psychological research: concep- Behavior and Human Performance 30, 109–131.
tual, strategic, and statistical considerations. Journal of Per- Grant A.M., Curtayne L. & Burton G. (2009) Executive coaching
sonality and Social Psychology 51, 1173–1182. enhances goal attainment, resilience and workplace well-being:
Bauer T.N., Erdogan B., Liden R.C. & Wayne S.J. (2006) A a randomized controlled study. The Journal of Positive Psy-
longitudinal study of the moderating role of extraversion: lea- chology 4, 396–407.
der-member exchange, performance, and turnover during new Hernandez S.R. (2005) The perfect storm: nursing in the twenty-
executive development. Journal of Applied Psychology 91, first century America. Health Affairs 24, 1372–1373.
298–310. Hom P.W. & Hulin C.L. (1981) A competitive test of the pre-
Blau P.M. (1964) Exchange and Power in Social Life. Wiley, New diction of reenlistment by several models. Journal of Applied
York, NY. Psychology 66, 23–39.

48 ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49
 LMX, job satisfaction and turnover intentions

Karsten M.A. (2010) Coaching: an effective leadership interven- Sagas M. & Cunningham G.B. (2004) Treatment discrimination
tion. Nursing Clinics of North America 45, 39–48. in college coaching: its prevalence and impact on the career
Lagace R.R., Castleberry S.B. & Ridnour R.E. (1993) An success of assistant basketball coaches. International Sports
exploratory salesforce study of the relationship between Journal 8, 76–88.
leader-member exchange and motivation, role stress, and Scandura T.A. & Graen G.B. (1984) Moderating effects of initial
manager evaluation. Journal of Applied Business Research 9, leader-member exchange status on the effects of a leadership
110–119. intervention. Journal of Applied Psychology 69, 428–436.
Laschinger H. (2007) The impact of leader-member exchange Shader K., Broome M.E., West M.E. & Nash M. (2001) Factors
quality, empowerment, and core self-evaluation on nurse influencing satisfaction and anticipated turnover for nurses in
managerÕs job satisfaction. Journal of Nursing Administration an academic medical center. Journal of Nursing Administration
37, 221–229. 31, 210–216.
Laschinger H.K., Finegan J. & Wilk P. (2009) Context matters: Simons S.R. (2008) Workplace bullying experienced by nurses
the impact of unit leadership and empowerment on nursesÕ newly licensed in Massachusetts and the relationship to inten-
organizational commitment. Journal of Nursing Administra- tion to leave the organization. Advances in Nursing Science 31,
tion 39 (5), 228–235. 48–59.
Le Blanc P.M., De Jong R.D., Geersing J., Furda J. & Komproe Smith G.P. (2000) Keeping good employees means having good
I.H. (1993) Leader member exchanges: distinction between two managers [WWW document]. Available at: [Link]
factors. European Work & Organizational Psychologist 3, [Link]/[Link]?id=32, accessed 21 June 2010.
297–309. Sparr J.L. & Sonnentag S. (2008) Fairness perceptions of super-
Lee T.H., Gerhart B., Weller I. & Trevor C.O. (2008) Under- visor feedback, LMX, and employee well-being at work.
standing voluntary turnover: path-specific job satisfaction ef- European Journal of Work and Organizational Psychology 17,
fects and the importance of unsolicited job offers. Academy of 198–225.
Management Journal 51, 651–671. Sparrow R.T. (1994) Empowerment in the hospital industry: an
Liden R.C. & Graen G. (1980) Generalizability of the vertical exploration of antecedents and outcomes. Hospitality Research
dyad linkage model of leadership. Academy of Management Journal 17, 51–73.
Journal 23, 451–465. Sparrow R.T. & Liden R.C. (2005) Two routes to influence:
Locke E.A. (1976) The nature and causes of job satisfaction. In integrating leader-member exchange and social network per-
Handbook of Industrial and Organizational Psychology (M.D. spectives. Administrative Science Quarterly 50, 505–535.
Dunnette ed.), pp. 1297–1349. Rand McNally, Chicago, IL. Spence-Laschinger H.K., Finegan J. & Shamian J. (2001) The
Matzler K. & Renzl B. (2006) The relationship between inter- impact of workplace empowerment, organizational trust on
personal trust, employee satisfaction, and employee loyalty. staff nursesÕ work satisfaction and organizational commitment.
Total Quality Management 17, 1261–1271. Health Care Management Review 26, 7–24.
McCarthy G., Tyrrell M. & Lehane E. (2007) Intention to ÔleaveÕ Tangirala S., Green S.G. & Ramanujam R. (2007) In the shadow
or ÔstayÕ in nursing. Journal of Nursing Management 15, 248– of the bossÕs boss: effects of supervisorsÕ upward exchange
255. relationships on employees. Journal of Applied Psychology 92,
McGowan B. (2001) Self-reported stress and its effects on nurses. 309–320.
Nursing Standard 15, 33–38. Tett R.P. & Meyer J.P. (1993) Job satisfaction, organizational
McKay P.F., Avery D.R., Tonidandel S., Morris M.A., Hernandez commitment, turnover intention, and turnover: path analyses
M. & Hebl M.R. (2007) Racial differences in employee based on meta-analytic findings. Personnel Psychology 46,
retention: are diversity climate perceptions the key? Personnel 259–293.
Psychology 60, 35–62. Thibault J.W. & Kelley H.H. (1959) The Social Psychology of
Paradis E. (2005) R for beginners [WWW document]. Available Groups. Wiley, New York, NY.
at: [Link] Trinkoff A., Geiger-Brown J., Brady B., Lipscomb J. & Muntaner
accessed 21 July 2010. C. (2006) How long and how much are nurses now working?
Podsakoff P.M., MacKenzie S.B. & Bommer W.H. (1996) Tran- Too long, too much, and without enough rest between shifts, a
formational leader behaviors and substitutes for leadership as study finds. American Journal of Nursing 106, 60–72.
determinants of employee satisfaction, commitment, trust, and Van Dick R., Christ O., Stellmacher J. et al. (2004) Should I stay or
organizational citizenship behaviors. Journal of Management go? Explaining intentions with organizational idemtification and
22, 259–298. job satisfaction. British Journal of Management 15, 351–360.
Preacher K.J. & Leonardelli G.J.. (2001) Calculation for the sobel Waldman J.D., Kelly F., Arora S. & Smith H.L. (2004) The
test [WWW document]. Available at: [Link] shocking cost of turnover in healthcare. Health Care Manage-
~preacher/sobel/[Link], accessed 9 September 2009. ment Review 29, 2–7.
Price J.L. & Mueller C.W. (1981) A causal model of turnover for Wayne S.J., Shore L.M. & Liden R.C. (1997) Perceived organi-
nurses. Academy of Management Journal 24, 543–565. zational support and leader-member exchange: a social
R Development Core Team (2010) R: A Language and Envi- exchange perspective. Academy of Management Journal 40,
ronment for Statistical Computing. R Foundation for Statistical 82–111.
Computing, Vienna, Austria. World Health Organization (2006) The World Health Report
Ramrup N. & Pacis M. (2008) The relationship between job 2006: Working Together for Health. WHO, Geneva.
satisfaction and job-related stress and how it influences Wright T.A. & Bonett D.G. (2007) Job satisfaction and psycho-
retention to leave oncology nursing. Oncology Nursing Forum logical well-being as nonadditive predictors of workplace
35, 538. turnover. Journal of Management 33, 141–160.

ª 2010 The Authors. Journal compilation ª 2010 Blackwell Publishing Ltd, Journal of Nursing Management, 19, 41–49 49