Role stress and personal resources in nursing: A cross-sectional study of burnout and engagement

Article (PDF Available)inInternational journal of nursing studies 48(4):479-89 · April 2011with1,000 Reads
DOI: 10.1016/j.ijnurstu.2010.08.004 · Source: PubMed
Abstract
The experience of role stress has been linked to burnout as an important job stressor, but the impact of this stressor in the context of engagement (characterized by vigor, dedication, and absorption) has not yet been sufficiently studied among nurses. Personal resources also appear to influence the process of burnout and engagement. This study examines the influence of role stress and personal resources (optimism, hardy personality and emotional competence) in nursing on burnout and engagement dimensions. Cross-sectional data from 508 nurses from general hospitals in Madrid (Spain) showed that both role stress and personal resources were related to burnout and engagement dimensions, although role stress was more closely related to nursing burnout, whereas personal resources were more closely related to nursing engagement. In addition, optimism as a personal resource, showed a moderator effect on exhaustion and the three dimensions of engagement. The study provides additional support about role stress as an important predictor of burnout and engagement in nursing, even after controlling for personal resources and socio-demographic variables.
Role stress and personal resources in nursing: A cross-sectional
study of burnout and engagement
Eva Garrosa
*
, Bernardo Moreno-Jime
´
nez, Alfredo Rodrı
´
guez-Mun
˜
oz,
Raquel Rodrı
´
guez-Carvajal
Faculty of Psychology, Universidad Auto
´
noma de Madrid, Ctra. de Colmenar km. 15, 28049 Madrid, Spain
What is already known about the topic?
Nursing burnout is more closely related to job
demands, whereas nursing engagement is more closely
related to job resources (physical, psychological, social,
or organizational aspects of the job that stimulate
personal growth and progress). Additionally, studies
about the analysis of the influence of personal
resources in nursing on burnout and engagement are
scarce.
Role stress is associated with negative consequences for
the nurses and the hospitals. For example, role stressors
predicted mental health in the workplace.
Optimism, hardy personality, and emotional competence
are relevant personal resources that affect the way in
which nurses interact with their working environment.
What this paper contributes
The paper adds to previous research by emphasizing that
role stress is an important factor in burnout and
engagement, even after controlling for personal resources.
The study has revealed the main effects of optimism,
hardy personality and emotional competence indicating
that nurses with these personal resources have less
burnout and more engagement. Additionally, nurses
with greater optimism reported lower levels of emo-
tional exhaustion and higher vigor, dedication and
absorption when experiencing higher role stress than
did less optimistic nurses.
International Journal of Nursing Studies xxx (2010) xxx–xxx
ARTICLE INFO
Article history:
Received 20 October 2009
Received in revised form 2 August 2010
Accepted 15 August 2010
Keywords:
Burnout
Emotional competence
Engagement
Hardy personality
Nursing
Optimism
Personal resources
Role stress
ABSTRACT
Background:
The experience of role stress has been linked to burnout as an important job
stressor, but the impact of this stressor in the context of engagement (characterized by
vigor, dedication, and absorption) has not yet been sufficiently studied among nurses.
Personal resources also appear to influence the process of burnout and engagement.
Objectives: This study examines the influence of role stress and personal resources
(optimism, hardy personality and emotional competence) in nursing on burnout and
engagement dimensions.
Design and results: Cross-sectional data from 508 nurses from general hospitals in Madrid
(Spain) showed that both role stress and personal resources were related to burnout and
engagement dimensions, although role stress was more closely related to nursing burnout,
whereas personal resources were more closely related to nursing engagement. In addition,
optimism as a personal resource, showed a moderator effect on exhaustion and the three
dimensions of engagement.
Conclusions: The study provides additional support about role stress as an important
predictor of burnout and engagement in nursing, even after controlling for personal
resources and socio-demographic variables.
ß 2010 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +34 92 4974548; fax: +34 91 4976133.
E-mail address: eva.garrosa@uam.es (E. Garrosa).
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Contents lists available at ScienceDirect
International Journal of Nursing Studies
journal homepage: www.elsevier.com/ijns
0020-7489/$ see front matter ß 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
The present study suggests the initial step to understand
the link between these variables, diminish nursing
burnout, and reinforce nursing engagement.
1. Introduction
The job demands-resources model (JD-R model;
Peterson et al., 2008; Bakker and Demerouti, 2007; Bakker
et al., 2005; Demerouti et al., 2001; Schaufeli and Bakker,
2004) specifies how employee well-being may be
produced by the interaction of job demands and job
resources. Job demands are the physical, social, or
organizational aspects of the job that require sustained
physical and/or psychological (i.e., cognitive or emo-
tional) effort on the part of the employee, and are
therefore associated with certain physiological and/or
psychological costs (Bakker et al., 2007). Job resources are
the physical, psychological, social, or organizational
aspects of the job t hat decrease the costs of the associated
physiological and psychological efforts of job demands
and they also stimulate personal growth, learning, and
development (Demerouti et al., 200 1). In general, job
demands and resources are negatively related because job
demands, such as high work pressure and emotionally
demanding interactions with patients, or clients, may
preclude the mobilization of job r esources (Bakker and
Demerouti, 2007; Demerouti et al., 2001).
The JD-R model (Demerouti et al., 2001) suggests two
psychologically interrelated processes. Specifically,
demanding aspects of work (i.e., role stress) lead to constant
overtaxing and, in the long run, to burnout (e.g., Bakker and
Demerouti,2007; Leeand Ashforth, 1996). The otherprocess
connects personal resources to engagement. In recent years,
burnout researchers have started to pay attention to the
conceptual opposite of burnout: engagement (Maslach
et al., 2001; Schaufeli et al., 2002). Job resources would
lead to engagement and positive outcomes (Schaufeli and
Bakker, 2004). For example, job resources would encourage
personal investment in the work and success of the
organization.
Engagement is characterized by energy, implication or
commitment, and efficiency, aspects directly opposite to
burnout (Schaufeli et al., 1996). Engagement has also been
defined as a positive motivational construct characterized
by vigor (high levels of energy and mental resilience while
working), dedication (refers to being strongly involved in
one’s work and experiencing a sense of meaning and pride),
and absorption (considered as being completely concen-
trated and happily immersed in one’s work) (Schaufeli et al.,
2002). In general, it refers to a cognitive-affective, persistent
state in time, which is not focused on an object or a specific
behavior. From the viewpoint of Maslach et al. (2001),
engagement is a perspective that emphasizes the person’s
relation with his or her work.
The results found until now using both processes,
burnout and engagement, have shown that burnout is
more closely related to job demands—overload, emotional
demands, etc. (Bakker et al., 2003, 2004), whereas
engagement is more closely related to job resources—
control, feedback, opportunities of learning, etc. (Hakanen
et al., 2006; Maslach et al., 2001).
Nevertheless, studies of the prediction of nursing
engagement are scanty (Schaufeli and Bakker, 2004;
Simpson, 2008). Studies about the analysis of the
influence of personal resources on engagement are still
scarcer (Hakanen et al., 2006; Langelaan et al., 2006;
Mauno et al., 2007; Xanthopoulou et al., 2007).
Consequently, there is a major need for studies on the
relationship of burnout and engagement with personal
resources and its antecedents.
2. Role stress
Research has revealed that job-related stressors (e.g.,
work overload, emotional demands, role ambiguity or
uncertainty, lack of autonomy) may have a profound impact
on burnout (for reviews, see
Lee and Ashforth, 1996;
Schaufeli and Enzmann,1998). Role theory (e.g., Jackson and
Schuler, 1985; Katz and Kahn, 1978) suggests that interrole
conflict and tension often result, as individuals find it
increasingly difficult to perform successfully each of their
roles, either due to limited resources (e.g., energy and time)
or to incompatibility of diverse roles (e.g., employee roles vs.
family roles). Specifically, role stress emerges from the
impact of the environment on an individual’s ability to fulfill
role expectations (Beehr and Glazer, 2005).
Some studies suggest that the experience of being
overworked is increasing in nursing (Chang et al., 2005;
Cartwright and Cooper, 1997; Jawahar et al., 2007; Garrosa
et al., 2008; Kelloway and Day, 2005; O
¨
rtqvist and Wincent,
2006). Issues of workload and workplace become increas-
ingly important in an environment in which working hours
and workdemandsare increasing(Kelloway and Day, 2005).
Workload associated with being overworked has been
found to have a negative impact across behavioral,
psychological, and physiological outcome domains (McVi-
cara, 2003). Role ambiguity has also been especially
associated with increased symptoms of stress (Day and
Livingstone, 2001).
Anchored in role theory, role stress is associated with
negative consequences for the individual and the organiza-
tion. An enormous amount of research documents the
impact of role stressors (for a review, see Beehr and Glazer,
2005; Chang et al., 2005; O
¨
rtqvist and Wincent, 2006). For
example, role stressors predicted mental health in the
workplace (Kelloway and Barling, 1991; Kelloway and Day,
2005). These effects included burnout (see Lee and Ashforth,
1996; O
¨
rtqvist and Wincent, 2006). Burnout is often cited as
a consequence of role stress, because experiences of stress
eventually deplete emotional resources (Posig and Kickul,
2003).In addition,role stress is a key to understanding levels
of personal accomplishment (Kelloway and Barling, 1991;
Peiro
´
et al., 2001). The relationship between depersonaliza-
tion and role stress is based on a coping response. When
individuals experience high role stress levels, they are likely
to distance themselves from work as a way of coping
(Golembiewski et al., 1996).
3. The job resources and personal resources
However, there is little previous empirical research on
antecedents of nursing engagement. Due to the fact that
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
2
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
the JD-R model has predominated in work engagement in
general research, current studies have concentrated more
on job and organizational resources. A recent study
confirmed that time demands, work-to-family conflict,
and job insecurity were each related to a separate
dimension of work engagement. Time demands predicted
high absorption, work-to-family conflict predicted low
vigor, and job insecurity predicted low dedication (Mauno
et al., 2007).
With regard to job resources, Salanova et al. (2005a)
found that training possibilities in organizations and job
autonomy (i.e., job control) were related to high work
engagement. In addition, Bakker and Geurts (2004) also
found that job autonomy, possibilities for professional
development, and feedback on performance at work
especially increased experiences of absorption at work.
Absorption, in turn, partially mediated the relationship
between job resources and a positive work-to-home
interface. Likewise, Bakker (2005) has shown that four
specific job resources—social support at work, supervisory
coaching, job autonomy, and performance feedback at
work—were associated with high engagement.
4. Optimism, hardy personality and emotional
competence as personal resources
Traditional research of personality and burnout has
focused on the Big Five personality dimensions (extrover-
sion, agreeableness, conscientiousness, neuroticism and
openness) (Bakker et al., 2006; Barrick and Mount, 1991;
Hurtz and Donovan, 2000), or on positive and negative
affectivity (Cooper, 2000; Payne, 2000; Spector et al.,
2000). Specifically, the results indicated that extroversion
significantly predicted diminished accomplishment, and
neuroticism significantly predicted exhaustion and deper-
sonalization. Other authors (Zellars et al., 2004; Varvel et
al., 2007) have extended the research effort on personality
resources and moods at work.
Recently, the emerging positive psychology is empha-
sizing the relative importance that positive psychological
strengths and capacities can have on human functioning
(Csikszentmihalyi, 1990; Seligman and Csikszemtmihalyi,
2000; Lent, 2004; Maslach et al., 2001; Peterson and
Seligman, 2004; Schaufeli and Bakker, 2004; Zellars et al.,
2004).
The positive psychology orientation offers an approach
to some positive aspects of personal resources, thus aiding
our comprehension of the processes of burnout and
engagement. Therefore, interventions aimed at reducing
risk for nursing burnout and improving nursing engage-
ment may be more effective if they include enhancing
nurses’ personalities rather than just decreasing environ-
mental stressors, for example by learning to take an
optimistic view, developing emotional competencies (i.e.,
discerning others’ emotions, verbal expression of emo-
tions, and empathy) and knowledge of hardy personality
skills (i.e., learning to see a challenge in the changing
situations) (Garrosa et al., 2008; Peterson and Seligman,
2004; Maddi, 2002). The positive psychology approach has
also an organizational application. Particularly, the affir-
mative attitudes that comprise positive psychology
correspond, at the organizational level, to cooperation,
credibility, creativity, and well-being (Maddi, 2002; Maddi
et al., 1999).
Positive psychology has listed many personal resources
(for a review, see Snyder and Lopez, 2002; Peterson and
Seligman, 2004), some of them linked to the study of
burnout, such as optimism (Chang et al., 2000; Grau et al.,
2005; Riolli and Savicki, 2003), self-esteem (Browning et
al., 2006; Janssen, 1999; Ma
¨
kikangas and Kinnunen, 2003;
Pruessner et al., 1999), hardy personality (Chang, 2003;
Michielsen et al., 2004), and self-efficacy (Cherniss, 1993;
Salanova et al., 2005b; Xanthopoulou et al., 2007). Still
other personal resources have been considered, such as
sense of coherence, emotional competence, flow, or
resilience, among others, but with less tradition in the
research on burnout.
There is no doubt that these positive constructs are
related, but there are not many studies of several of them
simultaneously and none of burnout and engagement
together. If certain aspects of the person as an individual
agent and control overlap (Smith et al., 2000), they should
be studied in interdependent combinations as well as
singly. From this perspective, optimism, hardy personality,
and emotional competence, taken together, seem to be
relevant personal resources that affect the way in which
nurses interact with their working environment (Scheier
and Carver, 1985; Kobasa, 1979; Saarni, 1999). These
personal resources have been identified as important
factors of appraisal, adaptive strategies and subjective
well-being.
Optimism has been described as an activity related to
goal attainment, positive expectations, and self-regulation
(Scheier and Carver, 1985). According to this definition,
self-regulators plan ahead to manage obstacles that might
otherwise thwart their goals. They also display confidence
that, despite difficult obstacles, they will still persist in
achieving their goals (Lebsack et al., 2004). Optimism can
be linked to higher workplace performance and lower
burnout (Chang et al., 2000; Grau et al., 2005; Riolli and
Savicki, 2003). Optimistic individuals are more likely to
make a plan of action for difficult situations (Strack et al.,
1987), are less likely to give up (Seligman, 1998), and have
a more positive outlook on stressful situations (Carver and
Gaines, 1987). There is also some evidence that optimism
can be positively associated with affective measures of
employee attitudes such as increased job satisfaction and
organizational commitment (Agho, 1993; Lum et al., 1998;
Segerstrom, 2007; Wanberg and Banas, 2000).
Hardy personality has been proposed by Kobasa (1979)
as a distinctive and active way of understanding a person’s
relation with others, with goals, and with problems (Maddi
et al., 2006; Kobassa-Ouellette and Di Placido, 2001). This
model of positive psychology is made up of three variables:
Commitment refers to engagement in life, control is related
to the recognitionof one’s influenceon events,and challenge
is orientation to change. Two mechanisms have been
proposed to explain the effect of hardiness: a more
optimistic perception of events (Alfred and Smith, 1989;
Wiebe, 1991) and the use of specific coping strategies
(Westman, 1990; Williams et al., 1992). According to Ford-
Gilboe and Cohen (2000), a hardy person’s transformational
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
3
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
way of coping involves either changing the stressful life
events themselves or thinking about them optimistically.
Another complementary action is the improvement of
personal health practices, a consequence of one’s beliefs
about one’s own health behaviors. Hardiness appears to
havea positive association with organizational effectiveness
(Ford-Gilboeand Cohen, 2000; Moreno-Jime
´
nezet al., 2006;
Garrosa, 2006; Pollock, 1986) and a negative relationship
with burnout. Findings provide partial support for the
hypothesis that greater hardiness is associated with less
stress and burnout (Duquette et al., 1995; Kobassa-
Ouellette and Di Placido, 2001; Michielsen et al., 2004;
Rich and Rich, 1987).
Emotional competence refers to diverse competence
skills necessary in social-emotional transactions (Saarni,
1999). In the workplace context, emotional competence
implies managing one’s emotions in accordance with one’s
moral dispositions and ethical values. Saarni adopts a
developmental perspective and states that, as people
acquire the skills of emotional competence, they display
three consequences in their behavior: the skill to manage
their own emotions with coping strategies, a sense of
subjective well-being facilitated by their capacity for
emotional self-efficacy, and, lastly, resilience or the ability
to recover rapidly after experiencing an adverse situation.
These consequences have been confirmed in personal,
social, and workplace contexts (Asher and Rose, 1997;
Compas et al., 1992; Crick and Dodge, 1994; Denham, 1999;
Garber et al., 1991; Parke et al., 1992; Wolchik and Sandler,
1997).
The aim of the present study is to examine the
interaction of the situational context of the current role
and personal resources in relation to burnout end
engagement dimensions. For this purpose, the effects
of role stress and three personal resources were
analyzed: optimism, hardy personality, and emotional
compet ence.
4.1. Present study
By investigating the potential role of personal resources
in the relation between role stress and nursing burnout
(emotional exhaustion, depersonalization, and lack of
personal accomplishment) and nursing engagement (ded-
ication, vigor, and absorption), we extend previous research
on burnout and engagement, and contribute information
about the positive impact of personal resources on role
stress, burnout, and engagement among nurses.
The present study tests the following hypotheses:
Hypothesis 1.
Role stress is positively related to dimen-
sions of nursing burnout (emotional exhaustion, deper-
sonalization and lack personal accomplishment) and
negatively related to nursing engagement (dedication,
vig or and absorption) when socio-demographic variables
and personal resources have been taken into account.
Hypothesis 2.
Personal resources are negatively related to
nursing burnout dimensions and positively related to nur-
sing engagement dimensions when socio-demographic
variables and role stress have been taken into account.
Hypothesis 3.
A high level of personal resources is asso-
ciated with less nursing burnout.
Hypothesis 4.
A high level of personal resources is asso-
ciated with more nursing engagement.
Hypothesis 5.
Personal resources have a moderating effect
on the relationship between role stress and the dimensions
of burnout and engagement, respectively.
5. Method
5.1. Participants and procedures
The study was approved by the Ethics Committee of the
Autonomous University of Madrid, CEI-21-446, by which
human subjects’ protection is ensured. The researcher
explained the purpose of the current study to nursing
directors of hospitals. Questionnaires were distributed to
potential participants with a cover page, explaining the
purpose of study and providing assurance of the anon-
ymous, voluntary and confidential nature of the responses.
Informed written consent was given by the responders. To
maintain confidentiality, the questionnaires were identified
with numbers. We followed this recommendation proposed
by Podsakoff et al. (2003) to reduce common source bias in
cross-sectional design with survey questionnaire.
A total of 508 nurses from four general hospitals in
Madrid (Spain) completed a self-report (65.5% of the
participants were permanent nurses, 34.5% temporary
nurses). Of the participants, 89.6% were female and the
majority (70.5%) spent more than 80% of their working
time interacting with patients. About 65% of the nurses
worked in the morning or on rotating shifts. The overall
response rate was 61%.
5.2. Measures
Table 1 shows the intercorrelations, means, standard
deviations, and reliabilities (Cronbach’s
a
) of the variables
included in the study. All scale reliabilities were satisfac-
tory.
5.2.1. Socio-demographics
Gender (male coded as 0 and female as 1), job status
(permanent nurses coded as 1, temporary-contract nurses
coded as 2), shift (morning coded as 1, afternoon coded as 2,
night coded as 3, and rotating coded as 4), percentage of
interaction with patients (more than 80% coded as 1,
between 60% and 40% coded as 2, between 40% and 20%
coded as 3, less than 20% coded as 4), and years of experience
(0–5 years coded as 1, >5–10 years coded as 2, >10 years
coded as 3) were included as control variables.
5.2.2. Role stress and burnout
Role stress and burnout were assessed using the
Nursing Burnout Scale (NBS) (Garrosa et al., 2008;
Moreno-Jime
´
nez et al., 2000a). This scale is a specific
measurement of the nursing burnout, with explicit
antecedents. Burnout (24 items) in this scale is consistent
with the three dimensions proposed by Maslach and
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
4
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
Jackson (1986), albeit the dimension of personal accom-
plishment has been replaced with its inverse or lack of
personal accomplishment, to facilitate the interpretation
of profiles and the calculation of a global index of burnout
(established from the average obtained by each nurse in
each of three dimensions).
Furthermore, the NBS has a global index of role stress,
established from the average obtained by each nurse in
each of two sub-scales. Workload: 14-items evaluate the
excess of quantitative and qualitative demands (e.g., ‘‘I
have to attend too many patients’’). And Role ambiguity:6-
items were used to measure the clarity of information that
nurses perceive in relation to the labor and organizational
role (e.g., ‘‘The orders I receive are vague and ambiguous’’).
Each item has a 4-point response scale ranging from 1
(totally disagree)to4(totally agree). The NBS has been
found to have adequate reliability and validity (Garrosa
et al., 2008; Moreno-Jime
´
nez et al., 2000a).
5.2.3. Engagement
Engagement was measured with the Utrecht Work
Engagement Scale (UWES) ( Schaufeli et al., 2002). The
engagement items reflect three underlying dimensions:
vigor (VI), dedication (DE), and absorption (AB). The
engagement items are scored from 0 (never to 6 always).
High scores on VI, DE, and AB indicate engagement. This
17-item instrument has been found to have adequate
reliability and validity (Schaufeli et al., 2002). Never-
theless, the reliability estimate on the AB from the sample
data collected was low in this sample of nurses (
a
= .65).
5.2.4. Optimism
The Revised Life Orientation Test (LOT-R) (Scheier et al.,
1994) is a 6-item measure (plus 4 filler items) of individual
differences in dispositional optimism–pessimism. Partici-
pants are asked to respond to these items using a 5-point
Likert-type scale ranging from 0 (strongly disagree)to4
(strongly agree). The LOT-R is a modified brief version of the
original LOT (Scheier and Carver, 1985) and has been found
to have a correlation of .95 with the latter (Scheier et al.,
199 4).
5.2.5. Hardy personality
The Hardy Personality Scale (HPS) (Garrosa et al., 2008;
Moreno-Jime
´
nez et al., 2000b) is a 17-item measure of
commitment, control, and challenge. Items were rated
from 1 (wholly agree)to4(wholly disagree). The HPS has
been found to have adequate reliability and validity.
Internal consistency was
a
= .89 for the general scale
(Garrosa et al., 2008; Moreno-Jime
´
nez et al., 2000b).
5.2.6. Emotional competence
Emotionalcompetence was assessed using theEmotional
Competence Scale (ECS) (Garrosa et al., 2005). This 12-item
scale (4 items for each dimension) assessed the following
dimensions: Discerning Others Emotions (DOE) (e.g., ‘‘I
really don’t know how my friends or people who are close to
me feel if they don’t tell me’’). Verbal expression of the
emotions (VE) was used to measure the skill to use a wide
and varied vocabulary to express one’s emotions (e.g., ‘‘I can
express my feelings verbally without difficulty’’). And
empathy (EM) is the ability to understand another’s feelings
and experience (e.g., ‘‘It is easy for me to understand how my
friendsand relatives feel’’). The emotional competence items
are scored from 1 (totally disagree)to5(totally agree). High
scores on DOE, VE, and EM indicate emotional competence.
This 12-item instrument has been found to have adequate
reliability and validity (Garrosa et al., 2005, 2008).
5.3. Methods of analysis
The main and interactive effects of role stress and
personal resources on burnout and engagement were tested
using hierarchical multiple regression analysis. Socio-
demographic characteristics were controlled and entered
in the first step. Role stress was entered in the second step,
and personal resources the third step, in order to investigate
their relative importance. The interaction terms of role
stress with each of the three personal resources were then
entered in the fourth step to test for the hypothesized
moderating effect of personal resources on the relation
between role stress and burnout, and role stress and
engagement. Following the procedures described by Aiken
and West (1991), the predictor variables were centered, that
is, the means of these variables were set to zero while the
standard deviations werekept intact. The data were checked
for multicollinearity using tolerance and the variance
inflation factor (VIF) (e.g., see Kleinbaum et al., 1988).
VIF-valuesgreaterthan10 and tolerance-values smallerthat
.10 may indicate multicollinearity. There were no signs of
Table 1
Means, standard deviations, and correlations of all variables.
Measures M SD12345678910
Role stress 2.52 .42 (.84)
Emotional competence 3.76 .46 .16
**
(.89)
Optimism 2.57 .69 .05 .24
**
(.78)
Hardy personality 3.03 .34 .23
**
.28
**
.17
**
(.89)
Emotional exhaustion 2.29 .57 .54
**
.21
**
.22
**
.45
**
(.89)
Depersonalization 1.75 .46 .30
**
.32
**
.16
**
.47
**
.52
**
(.74)
Lack personal accomplishment 1.73 .47 .39
**
.32
**
.23
**
.50
**
.62
**
.60
**
(.76)
Dedication 4.49 1.13 .31
**
.28
**
.21
**
.58
**
.54
**
.47
**
.41
**
(.84)
Vigor 4.03 1.20 .22
**
.25
**
.22
**
.50
**
.46
**
.40
**
.31
**
.72
**
(.77)
Absorption 4.08 1.03 .29
**
.26
**
.31
**
.47
**
.56
**
.36
**
.40
**
.75
**
.66
**
(.65)
Note: reliability coefficients (Cronbach’s
a
) in parentheses along main diagonal. Analyses based on n = 508 nurses.
**
p < .01.
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
5
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
multicollinearity in any of the six regression models. All
analyses were performed with the SPSS-program (SPSS,
1990a,b). Six hierarchical regression analyses were per-
formed, one for each outcome variable on burnout and
engagement. The results are presented in Tables 2 and 3,
respectively. As a final point, to better explore the
moderating effects, we plotted the interactions using the
standardized regression coefficients of the regression lines
for employees high (1 SD above the mean) and low (1 SD
below the mean) on the moderator variable (Aiken and
West, 1991). Graphical representations of these interactions
are shown in Figs. 1–4.
6. Results
In the analysis of emotional exhaustion, only job status
and shift, entered in Step 1, emerged as significant
predictors. Job status was negatively related to emotional
exhaustion in nursing; this relationship indicates that
permanent nurses were more emotionally exhausted than
temporary-contract nurses. Shift was negatively related to
emotional exhaustion; in this case, the nurses with
rotating shifts were more emotionally exhausted than
nurses working in the morning shift.
In Step 2, when role stress was entered, job status, years
of experiences and role stress accounted for 33% of the
variance of emotional exhaustion. Role stress was the more
significant variable and was positively related to emotional
exhaustion (
D
R
2
= .16, p < .001).
In Step 3, the three personal resources were entered,
increasing the amount of variance explained (
D
R
2
= .11,
p < .001), while the effect of job status, years of experience,
and role stress remained significant. Optimism and hardy
personality were negatively related to emotional exhaus-
tion.
Finally, when the three interaction terms were entered
in the last step (
D
R
2
= .01, p < .001), one of these was
significant, and the effect of socio-demographic variables
were no longer significant. The significant interaction
between role stress and optimism indicates that nurses
Table 3
Results of hierarchical regression analyses, standardized regression coefficients. Predictors of nursing engagement.
Vigor Dedication Absorption
Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4
Gender .02 .04 .02 .02 .04 .06 .05 .05 .06 .06 .07 .07
Job status .22
**
.16
*
.04 .09 .45
***
.40
***
.19
**
.12 .31
***
.28
***
.08 .04
Shift .15
**
.10
*
.08 .08 .11
**
.06 .03 .03 .12
*
.10
*
.07 .07
% Interaction with patients .07 .10
*
.09
*
.11
*
.06 .09
*
.09
*
.11
**
.02 .04 .04 .05
Years of experience .04 .04 .07 .09 .09 .09 .10 .14
*
.08 .08
*
.09 .11
Role stress (RS) .21
***
.17
***
.17
***
.20
***
.16
***
.18
***
.10 .07 .08
Emotional competence (EC) .06 .07 .09
*
.11
**
.07 .08
Optimism (O) .22
***
.22
***
.12
**
.13
**
.13
**
.13
**
Hardy personality (HP) .38
***
.38
***
.40
***
.40
***
.39
***
.39
***
EC RS .06 .06 .01
O RS .14
**
.20
***
.10
**
HP RS .03 .05 .01
R
2
(adjusted) .10 .14 .34 .35 .31 .34 .52 .55 .18 .19 .36 .36
D
R
2
.10 .04 .20 .01 .31 .03 .18 .03 .18 .01 .17 .01
*
p < .05.
**
p < .01.
***
p < .001.
Table 2
Results of hierarchical regression analyses, standardized regression coefficients. Predictors of nursing burnout.
Emotional exhaustion Depersonalization Lack of personal accomplishment
Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4
Gender .01 .02 .01 .01 .04 .06 .06 .06 .05 .07 .07 .07
Job status .43
***
.31
***
.16
*
.12 .38
***
.32
***
.14 .15
*
.35
***
.27
***
.06 .06
Shift .18
***
.08 .07 .07 .11
**
.06 .03 .03 .12
**
.06 .02 .03
% Interaction with patients .04 .02 .03 .04 .02 .02 .01 .01 .01 .06 .05 .05
Years of experience .14 .14
*
.13
*
.11 .04 .04 .05 .06 .10 .10 .09 .09
Role stress (RS) .44
***
.41
**
.42
**
.22
***
.18
***
.17
***
.31
***
.27
***
.25
***
Emotional competence (EC) .00 .01 .18
***
.18
***
.14
**
.14
**
Optimism (O) .13
**
.14
***
.01 .01 .11
**
.11
**
Hardy personality (HP) .31
***
.30
***
.37
***
.37
***
.39
***
.39
***
EC RS .06 .01 .06
O RS .12
**
.03 .01
HP RS .03 .04 .03
R
2
(adjusted) .17 .33 .43 .44 .14 .18 .33 .33 .11 .18 .37 .37
D
R
2
.17 .16 .11 .01 .14 .04 .16 .00 .11 .07 .18 .00
*
p < .05.
**
p < .01.
***
p < .001.
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
6
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
with greater optimism reported lower levels of emotional
exhaustion when experiencing higher role stress than did
less optimistic nurses (see Fig. 1). In total, the model
variables accounted for 44% of the variation of emotional
exhaustion.
In the results for depersonalization, job status and shift
again emerged as significant predictors (R
2
= .14, p < .001).
In Step 2, when role stress was entered, the proportion of
explained variance increased (
D
R
2
= .04, p < .001), while
the effects of job status remained significant. In Step 3,
emotional competence and hardy personality were nega-
tively related to depersonalization, while optimism did not
emerge as a significant predictor. However, the effect of
socio-demographic variables was no longer significant and
none of interaction terms had any impact on depersona-
lization.
Lack of personal accomplishment displayed a some-
what similar pattern. In Step 1, job status and shift were
significant predictors. In Step 2, when role stress was
entered, the proportion of explained variance increased
(
D
R
2
= .07, p < .001) and only the effect of job status
remained significant. Emotional competence, optimism,
and hardy personality were negatively related to lack of
personal accomplishment. Only significant role stress
coefficients remained in Step 3, and large changes in
magnitude appeared, as the personal resources variables
were entered (
D
R
2
= .18, p < .001). Once more, none of
these interaction terms had any impact on the criterion
variable.
The results for the three engagement variables showed
some significant differences. Job status and shift were
entered along with vigor in Step 1. Job status was positively
related to vigor, and shift showed a negative relation with
the criterion variable. In Step 2, role stress was negatively
related to vigor, and the effects of job status, shift, and
percentage in interaction with the patients continued
significant. These relationships indicate that nurses with
rotating shift and with high percentages of interaction
with patients have lower levels of vigor. In Step 3,
optimism and hardy personality showed a positive relation
with vigor, and there was a sharp increase in the amount of
variance explained (
D
R
2
= .20, p < .001), while the effects
of the percentage of interaction with patients and role
stress remained significant. Then, the significant interac-
tion emotional competence and hardy personality had no
impact on vigor. Only was significant interaction between
role stress and optimism indicates that nurses with greater
optimism reported higher levels of vigor when experien-
cing higher role stress than did less optimistic nurses (see
Fig. 2). In sum, 35% of the variation of vigor was accounted
for by the model variables.
Job status and shift were significant predictors of the
significant variables for dedication as a criterion, account-
ing for 31% of the variance in Step 1. When role stress was
entered in Step 2, the effect of shift was no longer
significant. Job status, percentage of interaction with
patients, and role stress were significant predictors in
Step 2. When the positive variables were entered,
optimism and hardy personality emerged as significant
predictors (
D
R
2
= .18, p < .001); both variables were
positively related to dedication. When the three interac-
tion terms were entered in the last step (
D
R
2
= .03,
p < .001), the interaction between optimism and role
stress was significant (see Fig. 3), and the effect of shift was
no longer significant. This model accounted for 55% of the
variation in dedication.
Finally, in the results for absorption, job status, and
shift emerged as significant predictors in Step 1. In Step 2,
Fig. 1. Interaction effects of role stress and optimism in predicting
emotional exhaustion.
Fig. 2. Interaction effects of role stress and optimism in predicting vigor.
Fig. 3. Interaction effects of role stress and optimism in predicting
dedication.
Fig. 4. Interaction effects of role stress and optimism in predicting
absorption.
E. Garrosa et al. / International Journal of Nursing Studies xxx (2010) xxx–xxx
7
G Model
NS-1724; No. of Pages 11
Please cite this article in press as: Garrosa, E., et al., Role stress and personal resources in nursing: A cross-sectional
?study of burnout and engagement. Int. J. Nurs. Stud. (2010), doi:10.1016/j.ijnurstu.2010.08.004
Trial Version
www.nuance.com
role stress was not a significant predictor of absorption. In
the Step 3, optimism and hardy personality were positively
related to absorption, and the effects of socio-demo-
graphics variables were no longer significant. The inter-
action of optimism and role stress affected absorption (see
Fig. 4).
7. Discussion
Basedon the JD-R model, the present study examined the
relationship between role stress and personal resources and
burnout and engagement among nurses. The positive
relation between role stress and emotional exhaustion,
depersonalization, and lack of personal accomplishment is
well documented (Lee and Ashforth, 1996; Schaufeli and
Enzmann, 1998). However, the present study found a
significant negative association between role stress and
each of three dimensions of nursing engagement. Once
again, depersonalization and role stress were correlated.
However, this relationship is not necessarily a function of
using distancing as a coping mechanism. Depersonalization
may be a counterproductive work behavior that causes role
stress. This result would be interesting to consider in a
longitudinal study.
Personal resources had important effects, specifically
hardy personality was significant in all analyses of burnout
and engagement. Optimism had a principal effect on all
dimensions of engagement, on emotional exhaustion, and
on lack of personal accomplishment, but not on deperso-
nalization. Finally, emotional competence was a significant
predictor of depersonalization, lack of personal accomplish-
ment, and dedication. Regarding the moderator role of
personal resources, only optimism had significant effects on
the three dimensions of engagement and exhaustion. This
result mayindicate that moreoptimisticnursesreport lower
levels of emotional exhaustion and higher vigor, dedication,
and absorption when experiencing higher role stress than
do less optimistic nurses. This result may be attributed to
optimistic people’s more positive outlook