1. No category

HEC Montréal
Affiliée à l'Université de Montréal
Le processus d'épuisement professionnel
chez les travailleurs du secteur de la santé
et ses conséquences sur la décision de se présenter au travail,
sur la performance et sur l'intention de quitter une profession
par
Geneviève Jourdain
Service de l'enseignement de la gestion des ressources humaines
HEC Montréal
Thèse présentée en vue de l'obtention du grade de
Philosophiæ Doctor (Ph. D.) en administration
Spécialisation gestion des ressources humaines
Février 2010
© Geneviève Jourdain, 2010
HEC Montréal
Affiliée à l'Université de Montréal
Cette thèse intitulée:
Le processus d'épuisement professionnel chez les travailleurs du secteur de la santé
et ses conséquences sur la décision de se présenter au travail,
sur la performance et sur l'intention de quitter une profession
présentée par:
Geneviève Jourdain
a été évaluée par un jury composé des personnes suivantes:
------------------------------------------------------------------------------------Caroline Aubé, Ph.D., présidente-rapportrice, HEC Montréal
-----------------------------------------------------------------------------------------------------------
Denis Chênevert, D.Sc. directeur de recherche, HEC Montréal
------------------------------------------------------------------------------------------------------------
Michel Tremblay, D.Sc., membre du jury, HEC Montréal
-------------------------------------------------------------------------------------------------------------
Steve Harvey, Ph.D., examinateur externe, Université Bishop's
--------------------------------------------------------------------------------------------------------------
François Bellavance, Ph.D., représentant du doyen de la FES, HEC Montréal
ii
Résumé
L'objectif général de la présente thèse est d'approfondir la compréhension du
processus d'épuisement professionnel chez les travailleurs du secteur de la santé et de
ses conséquences sur la décision de se présenter au travail, sur la performance au
travail et sur l'intention de quitter une profession. S'appuyant sur le modèle
demandes-ressources de l'épuisement professionnel (Demerouti, Bakker, Nachreiner,
& Schaufeli, 2001), la thèse rapporte les résultats de trois projets de recherche menés
auprès de travailleurs œuvrant au sein d'établissements rattachés au réseau canadien
public de la santé. Les données sont corrélationnelles et ont été collectées à l’aide
d’un questionnaire pré-testé et auto-administré, sauf pour le deuxième projet de
recherche pour lequel nous avons utilisé des données informatiques sur les absences
passées et futures compilées par l’établissement de santé.
Dans le premier article, nous proposons une nouvelle taxonomie de facteurs
de stress reliés à la vie au travail qui établit une distinction entre, d'une part, les
demandes et les ressources externes situées dans l'environnement de travail
comportant trois facettes (i.e. aspects liés à la tâche/rôle, aux interactions sociales ou
à l'organisation) et, d'autre part, les demandes et les ressources internes situées à
l'intérieur de l'individu mais induites par l'environnement de travail et associées à ses
trois facettes. Les relations postulées entre, d'une part, les demandes externes perçues
et les demandes internes induites associées à une même facette de l'environnement et,
d'autre part, entre les ressources externes perçues et les ressources internes induites
associées à une même facette de l'environnement ont toutes été confirmées. De plus,
les résultats tendent à supporter de manière générale notre hypothèse à l'effet que les
relations entre les facteurs de stress d'un même type, i.e. soit des demandes ou des
ressources, associés à une même facette de l'environnement sont plus fortes que les
relations entre les facteurs de stress d'un même type associés à des facettes différentes
de l'environnement. Quant au présumé rôle médiateur joué par les demandes et les
ressources internes induites par l'environnement dans les relations entre les facteurs
de stress externes perçus et l'épuisement professionnel, six des huit tests effectués ont
iii
confirmé la présence d'un effet de médiation. Enfin, les résultats suggèrent que les
demandes liées à la tâche ou au rôle et les ressources liées à l'organisation pourraient
être les facteurs contribuant le plus au développement de l'épuisement professionnel
chez les travailleurs du milieu hospitalier.
Dans le deuxième article, nous investiguons les relations entre les symptômes
d'épuisement professionnel chez les employés qui se présentent au travail, la
performance globale au travail et le présentéisme/absentéisme futur. Aussi, nous
examinons le rôle modérateur de la vision organisationnelle et des valeurs
d'excellence dans les relations entre les symptômes d'épuisement professionnel chez
les employés qui se présentent au travail et les comportements futurs de
présentéisme/absentéisme. Notre étude appuie fortement la proposition à l'effet
qu'une faible performance globale au travail, qui dit-on signalerait le présentéisme,
est un précurseur de l'absentéisme pour cause de santé. Aussi, les résultats révèlent
que l'articulation de la vision organisationnelle et la promotion de la qualité et de
l'innovation, telles que perçues par les employés en processus d'épuisement
professionnel, les incitent à choisir le présentéisme plutôt que l'absentéisme pour
cause de santé au cours de la prochaine année.
Enfin, dans le troisième article, nous étudions les relations entre les facteurs
de stress reliés au travail et à l'environnement social du personnel infirmier,
l'épuisement professionnel et l'intention de quitter la profession infirmière. Les
résultats de notre étude suggèrent que l'épuisement professionnel joue un rôle
important dans la décision du personnel infirmier de quitter sa profession, de part son
impact sur les troubles psychosomatiques et l'engagement professionnel. Afin de
retenir le personnel infirmier au sein de la profession, les résultats indiquent que ses
tâches et son rôle devraient être restructurés afin de réduire la charge de travail et
accroître le sens du travail.
iv
MOTS-CLÉS: Stress, Demandes, Ressources, Performance in-rôle, Performance
extra-rôle,
Présentéisme,
Absentéisme,
organisationnelles, Engagement professionnel
Vision
organisationnelle,
Valeurs
v
Abstract
The general aim of the present thesis is to gain a greater understanding of the
process of burning-out among health care workers and its consequences on the
decision to attend work, on job performance, and on the intention to leave a
profession. Based on the job demands-resources (JD-R) model of burnout
(Demerouti, Bakker, Nachreiner, & Schaufeli, 2001), the thesis reports the results of
three research projects carried out among workers of health care establishments of the
Canadian public health care sector. Data is cross-sectional and was collected with a
self-administrated and pre-tested questionnaire, except for the second research project
for which we used register data on previous and future sickness absences.
In the first article, we propose a new taxonomy of job stressors which
distinguishes between external demands and resources originating in the work
environment, which has three facets (i.e. aspects linked to the task/role, to social
interactions and to the organization), from internal demands and resources originating
in the person, but induced by the work environment and associated with its three
facets. The hypothesized relationships between perceived external and work-induced
internal demands associated with the same work environment facet, and between
perceived external and work-induced internal resources associated with the same
work environment facet were all confirmed. In addition, our presumption that
relationships between stressors of the same type, either demands or resources,
associated with the same work environment facet would be stronger than
relationships between stressors of the same type associated with different work
environment facets was generally supported. Regarding the postulated mediation
effects of work-induced internal stressors in the relationships between perceived
external stressors and the two burnout dimensions, six tests out of eight confirmed the
presence of a mediation effect. Finally, the results suggest that demands linked to the
task or role and resources linked to the organization may be the most important
contributors to the development of burnout among hospital workers.
vi
In the second article, we investigate the relationships between burnout
symptoms among employees who choose to bring themselves to work despite their
ill-health, an act known as presenteeism, overall job performance and future
presenteeism/absenteeism. Also, we examine the moderating role of organizational
vision and excellence values in the relationships between burnout symptoms among
working employees and future presenteeism/absenteeism. Our study provides strong
support for the proposition that low job performance said to signal presenteeism is a
precursor of sickness absenteeism. Also, the results reveal that the articulation of an
organizational vision and the promotion of quality and innovation, as perceived by
burning-out employees, prompt them to choose presenteeism over sickness
absenteeism in the following year.
Lastly, in the third article, we look at the relationships between stress factors
related to nurses’ work and social environment, burnout and intention to leave the
nursing profession. Results from our study suggest that burnout plays an important
role in nurses' decision to leave their profession, through its impact on psychosomatic
complaints and professional commitment. In order to retain nurses within the
profession, results indicate that nurses’ tasks and role should be restructured to reduce
work overload and increase the meaning of their work.
KEYWORDS: Stress, Demands, Resources, In-role performance, Extra-role
performance, Presenteeism, Absenteeism, Organizational vision, Organizational
values, Professional commitment
vii
Table des matières
LISTE DES TABLEAUX ............................................................................................ X
LISTE DES FIGURES ............................................................................................... XI
LISTE DES SIGLES ET ABREVIATIONS............................................................. XII
REMERCIEMENTS ................................................................................................ XIII
CHAPITRE 1: INTRODUCTION GÉNÉRALE .......................................................... 1
CHAPITRE 2: CADRE THÉORIQUE, OBJECTIFS DE RECHERCHE ET
APERÇU DES MODÈLES DE RECHERCHE.................................... 5
2.1
2.2
2.2.1
2.2.2
2.2.3
CADRE THEORIQUE ................................................................................. 5
OBJECTIFS DE RECHERCHE ET APERÇU DES MODÈLES DE RECHERCHE .... 8
Premier article ...................................................................................... 8
Deuxième article.................................................................................... 9
Troisième article.................................................................................. 10
CHAPITRE 3: MÉTHODOLOGIE ............................................................................ 15
CHAPITRE 4: PREMIER ARTICLE ......................................................................... 16
4.1
INTRODUCTION ..................................................................................... 18
4.2
THEORETICAL FRAMEWORK AND HYPOTHESES.................................... 19
4.2.2
A new taxonomy of demands and resources ........................................ 23
4.2.2.1 Determinants of burnout .................................................................. 23
4.2.2.2 Interrelations between demands and between resources ................. 27
4.2.3
Interrelated dual processes involved in the development of burnout .. 33
4.2.3.1 Energetic process ................................................................................ 34
4.2.3.2 Motivational process ........................................................................... 36
4.3
METHODOLOGY .................................................................................... 39
4.3.1
Participants ......................................................................................... 39
4.3.2
Measures ............................................................................................. 39
4.3.3
Analyses ............................................................................................... 42
4.4
RESULTS ............................................................................................... 43
4.4.1
Preliminary analysis............................................................................ 43
4.4.2
Measurement model ............................................................................ 44
4.4.3
Structural model .................................................................................. 45
4.5
DISCUSSION .......................................................................................... 54
4.5.1
Interpretation of results ....................................................................... 55
4.5.2
Study limitations .................................................................................. 59
4.5.3
Practical implications ......................................................................... 61
4.5.4
Research avenues ................................................................................ 62
CHAPITRE 5: DEUXIÈME ARTICLE...................................................................... 64
5.1
5.2
INTRODUCTION ..................................................................................... 65
THEORETICAL FRAMEWORK AND HYPOTHESES.................................... 68
viii
5.2.1
The job demands-resources model of burnout .................................... 69
5.2.2
Presenteeism as a precursor of sickness absenteeism......................... 71
5.2.2.1 The concept of presenteeism ........................................................... 71
5.2.2.2 Burnout and job performance .......................................................... 72
5.2.2.3 Burnout and absenteeism................................................................. 77
5.2.2.4 Job performance and absenteeism ................................................... 80
5.2.3
The boost effect of organizational vision and excellence values ......... 82
5.3
METHODOLOGY .................................................................................... 88
5.3.1
Participants ......................................................................................... 88
5.3.2
Measures ............................................................................................. 89
5.3.3
Analyses ............................................................................................... 91
5.4
RESULTS ............................................................................................... 92
5.4.1
Preliminary analysis............................................................................ 92
5.4.2
Hypothesis testing................................................................................ 93
5.5
DISCUSSION ........................................................................................ 101
5.5.1
Interpretation of results ..................................................................... 102
5.5.2
Study limitations ................................................................................ 106
5.5.3
Practical implications ....................................................................... 108
5.5.4
Research avenues .............................................................................. 110
CHAPITRE 6: TROISIÈME ARTICLE ................................................................... 113
6.1
INTRODUCTION ................................................................................... 114
6.2
THEORETICAL FRAMEWORK AND HYPOTHESES.................................. 116
6.2.1
Definition of Burnout......................................................................... 116
6.2.2
The Job Demands-Resources Model ................................................. 117
6.2.2.1 Demands and resources ................................................................. 120
6.2.2.2 Energetic process ........................................................................... 121
6.2.2.3 Motivational process ..................................................................... 123
6.2.2.4 Professional commitment .............................................................. 126
6.2.2.5 Intention to leave the profession ................................................... 127
6.3
METHODOLOGY .................................................................................. 129
6.3.1
Participants ....................................................................................... 129
6.3.2
Measures ........................................................................................... 130
6.3.3
Analyses ............................................................................................. 133
6.4
RESULTS ............................................................................................. 134
6.4.1
Preliminary analysis.......................................................................... 134
6.4.2
Development Phase of the Model ...................................................... 135
6.4.2.1 Measurement model ...................................................................... 135
6.4.2.2 Structural model ............................................................................ 136
6.4.3
Confirmatory Phase .......................................................................... 138
6.4.3.1 Structural model ............................................................................ 138
6.4.3.2 Additional analyses ....................................................................... 143
6.5
DISCUSSION ........................................................................................ 143
6.5.1
Interpretation of results ..................................................................... 144
6.5.2
Study limitations ................................................................................ 149
6.5.3
Practical implications ....................................................................... 150
6.5.4
Research avenues .............................................................................. 152
ix
CHAPITRE 7: DISCUSSION GÉNÉRALE ............................................................. 153
CHAPITRE 8: CONCLUSION GÉNÉRALE .......................................................... 163
RÉFÉRENCES .......................................................................................................... 165
x
Liste des tableaux
CHAPITRE 4: PREMIER ARTICLE
Table I:
Statistiques descriptives et corrélations...............................................46
Table II:
Relations entre les facteurs de stress de même type situés à des niveaux
non équivalents.....................................................................................50
CHAPITRE 5: DEUXIÈME ARTICLE
Table I:
Statistiques descriptives et corrélations...............................................94
Table II:
Résultats des analyses hiérarchiques de régression multiple pour les
problèmes de performance in-rôle et la performance extra-rôle/OCB-O
réduite...................................................................................................95
Table III:
Effet principal des dimensions clés de l'épuisement professionnel sur
les comportements d'absentéisme futurs pour cause de maladie.........97
Table IV:
Effet principal des types de performance au travail sur les
comportements d'absentéisme futurs pour cause de maladie..............98
Table V:
Effet d'interaction entre les dimensions clés de l'épuisement
professionnel et l'idéalisme organisationnel sur les comportements
d'absentéisme futurs pour cause de maladie......................................100
CHAPITRE 6: TROISIÈME ARTICLE
Table I:
Statistiques descriptives et corrélations.............................................139
xi
Liste des figures
CHAPITRE 4: PREMIER ARTICLE
Figure 1:
Modèle de recherche............................................................................25
Figure 2:
Résultats du test du modèle structurel..................................................48
CHAPITRE 5: DEUXIÈME ARTICLE
Figure 1:
Modèle de recherche............................................................................69
CHAPITRE 6: TROISIÈME ARTICLE
Figure 1:
Modèle de recherche..........................................................................118
Figure 2:
Résultats du test du modèle structurel................................................140
xii
Liste des sigles et abréviations
cf.
Comparer ou consulter
CFA
Confirmatory factor analysis
CFI
Comparative fit index
COR theory Conservation of resources theory
e.g.
Par exemple
Exp(B)
Exponent of the logistic coefficient
i.e.
C'est-à-dire
JD-R model Job demands-resources model
MBI-GS
Maslach Burnout Inventory - General Survey
MBI-HSS
Maslach Burnout Inventory - Human Services Survey
MLQ
Multifactor Leadership Questionnaire
Modèle DR
Modèle demandes-ressources
NFI
Normed fit index
OCB-I
Organizational citizenship behavior that immediately benefit specific
individuals
OCB-O
Organizational citizenship behavior that directly benefit the
organization
OCBs
Organizational citizenship behaviors
PE
Psychological empowerment
RMSEA
Root mean square error of approximation
RN
Registered nurse
SEM
Structural equation modeling
VIF
Variance inflation factor
xiii
Remerciements
En premier lieu, je remercie sincèrement mon directeur de recherche, le
Professeur Denis Chênevert, lui qui m'a tant encouragée à poursuivre des études
doctorales. Sa disponibilité, ses conseils, ses encouragements sur le plan personnel et
professionnel, son sens de l'humour et la confiance qu'il m'a toujours témoignée
m'ont permis de mener à bien cette thèse. Je lui suis reconnaissante de l'opportunité
qu'il m'a donnée de développer mes habiletés de chercheuse en me faisant participer
à ses travaux de recherche et de son soutien financier.
J'adresse ensuite mes remerciements aux membres du jury qui, par leurs
commentaires, m'ont permis d'enrichir cette thèse.
J'ai eu la chance d'être appuyée financièrement par divers organismes qui
m'ont permis de réaliser cette thèse dans des conditions idéales. Je remercie la
direction administrative du programme de Ph. D. de HEC Montréal, le Conseil de
recherches en sciences humaines, la Fondation J. Armand Bombardier et la Faculté
des études supérieures et postdoctorales de l’Université de Montréal.
J'exprime enfin ma profonde gratitude à mon conjoint, à ma mère et à ma
sœur pour leurs encouragements et le vif intérêt qu'ils ont porté à mon travail de
doctorante.
CHAPITRE 1: INTRODUCTION GÉNÉRALE
“When we fail to create internally motivating jobs for workers or when
we overload them with job stressors, when we treat workers as automata
or supermen rather than as human beings, the inevitable result is that the
quality of care for clients also becomes less humane ” (Eisenstat &
Felner, 1983).
On s'est d'abord intéressé à l'épuisement professionnel en tant que problème
social au début des années 70. Parmi les facteurs pouvant expliquer l'augmentation du
stress et de l'épuisement professionnel, Schaufeli et Enzmann (1998) notent la
croissance du secteur des services en général et des services sociaux en particulier
ainsi qu'un changement au niveau du contrat psychologique. Au cours des dernières
années, ils soutiennent que la relation d'échange réciproque entre l'employé et
l'organisation est devenue moins favorable à l'égard des employés: leur contribution
excède de manière considérable ce qu'ils reçoivent en retour de la part de leur
organisation.
Dans le secteur des services sociaux, les multiples réformes des systèmes de
santé menées dans plusieurs pays au cours de la dernière décennie ont été implantées
en dépit de leurs effets négatifs sur les ressources humaines (cf. Bourbonnais et al.,
1999; Greenglass & Burke, 2002; Shannon et al., 2001; White, 1997; Woodward et
al., 1999). Les stratégies visant à réduire les coûts et à améliorer l'efficacité à court
terme ont entraîné des problèmes tels qu'une surcharge de travail (cf. Adams,
Lugsden, Chase, Arber, & Bond, 2000) qui ont certes contribué à accroître les
perceptions des travailleurs du secteur de la santé d'une violation du contrat
psychologique. Les études démontrent qu'un déséquilibre entre les incitatifs et les
Chapitre 1: Introduction
2
résultats peut engendrer une détresse considérable et mener à l'épuisement
professionnel (e.g., Hendrix & Spencer, 1989; Schaufeli, Van Dierendonck, & Van
Gorp, 1996; Van Horn, Schaufeli, & Enzmann, 1999). De plus, les conditions de
travail actuelles dans les systèmes de santé ont non seulement un impact négatif sur la
santé des travailleurs mais aussi sur celle des patients (cf. Aiken, Clarke, & Sloane,
2002; Argentero, Dell’Olivo, & Ferretti, 2008; Garman, Corrigan, & Morris, 2002;
Halbesleben, Wakefield, Wakefield, & Cooper, 2008; Shanafelt, Bradley, Wipf, &
Back, 2002; Shirom, Nirel, & Vinokur, 2006).
La relation déséquilibrée entre l'employé et son organisation est un postulat
central du modèle demandes-ressources (DR) de l'épuisement professionnel
(Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Le modèle DR épouse la
perspective théorique dominante, mise de l'avant par Lazarus et Folkman (1984), qui
conçoit l'épuisement professionnel comme un état dysfonctionnel persistant résultant
d'une exposition prolongée au stress chronique, c'est-à-dire à une situation où la
personne se sent confrontée, de manière incessante, à des demandes élevées et des
ressources insuffisantes liées au travail comme tel et au contexte d'exécution du
travail (Buunk, de Jonge, Ybema, & de Wolff, 1998; Cooper, Dewe, & O'Driscoll,
2001; Leiter & Maslach, 2004).
Comme la recherche sur le modèle DR est relativement récente, l'objectif
général de la présente thèse est d'approfondir la compréhension du processus
d'épuisement professionnel chez les travailleurs du secteur de la santé et de ses
conséquences sur la décision de se présenter au travail, sur la performance au travail
et sur l'intention de quitter une profession. La thèse rapporte les résultats de trois
Chapitre 1: Introduction
3
projets de recherche menés auprès de travailleurs œuvrant au sein d'établissements
rattachés au réseau canadien public de la santé.
Dans le premier article, nous cherchons à clarifier comment les différents
facteurs de stress évoqués dans la littérature sont interreliés et contribuent au
développement de l'épuisement professionnel. Nous proposons une nouvelle
taxonomie de facteurs de stress associés à la vie au travail qui établit une distinction
entre, d'une part, les demandes et les ressources externes situées dans l'environnement
de travail comportant trois facettes (i.e. aspects liés à la tâche/rôle, aux interactions
sociales ou à l'organisation) et, d'autre part, les demandes et les ressources internes
situées à l'intérieur de l'individu mais induites par l'environnement de travail et
associées à ses trois facettes.
Dans le deuxième article, nous investiguons les relations entre les symptômes
d'épuisement professionnel chez les employés qui se présentent au travail, la
performance globale au travail et le présentéisme/absentéisme futur. Aussi, nous
examinons le rôle modérateur de la vision organisationnelle et des valeurs
d'excellence dans les relations entre les symptômes d'épuisement professionnel chez
les employés qui se présentent au travail et les comportements futurs de
présentéisme/absentéisme.
Enfin, dans le troisième article, considérant que les établissements de santé à
travers le monde font actuellement face à une pénurie de personnel infirmier
(International Council of Nurses & Florence Nightingale International Foundation,
2006), nous étudions les relations entre les facteurs de stress reliés au travail et à
Chapitre 1: Introduction
4
l'environnement social du personnel infirmier, l'épuisement professionnel et
l'intention de quitter la profession infirmière.
En résumé, le premier article examine plus spécifiquement les déterminants de
l'épuisement professionnel alors que le deuxième et le troisième articles traitent
davantage des conséquences de l'épuisement professionnel pour le travailleur,
l'organisation et la société.
CHAPITRE 2: CADRE THÉORIQUE, OBJECTIFS DE
RECHERCHE ET APERÇU DES MODÈLES DE RECHERCHE
2.1
Cadre théorique
La présente thèse s'appuie sur le modèle demandes-ressources (DR) de
l'épuisement professionnel (Demerouti et al., 2001). Ce modèle a été développé sur la
base d'un certain nombre de théories: le modèle de régulation de la performance par
contrôle compensatoire (Hockey, 1997), la théorie de conservation des ressources
(Hobfoll & Freedy, 1993), la théorie de fixation des objectifs (Locke & Latham,
1990) et la théorie de l'auto-détermination (Ryan & Deci, 2004).
Le modèle DR épouse la perspective théorique dominante mise de l'avant par
Lazarus et Folkman (1984). Selon cette perspective, l'épuisement professionnel est un
état dysfonctionnel persistant qui résulte d'une exposition prolongée au stress
chronique, c'est-à-dire à une situation où la personne se sent confrontée, de manière
incessante, à des demandes élevées et des ressources insuffisantes liées au travail
comme tel et au contexte d'exécution du travail (Buunk et al., 1998; Cooper et al.,
2001; Leiter & Maslach, 2004). Les demandes s’appliquent aux aspects
psychologiques, sociaux, organisationnels et physiques qui requièrent un effort de
l’employé et entraînent des coûts physiologiques et/ou psychologiques pour ce
dernier (Bakker, Demerouti, & Verbeke, 2004). Les ressources s’appliquent aux
aspects psychologiques, sociaux, organisationnels et physiques qui ont pour but de
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
6
faciliter l’atteinte des objectifs, de stimuler la croissance personnelle ou de réduire les
demandes (Bakker et al., 2004).
Les théoriciens qui ont développé le modèle DR ont adopté une conception
bidimensionnelle de l’épuisement professionnel inspirée par les travaux de Maslach
et Jackson (1996). L'épuisement réfère au sentiment d’être émotionnellement vidé et
épuisé sur les plans physique et cognitif par son travail. Le cynisme (ou
désengagement vis-à-vis du travail) reflète de l’indifférence ou une distanciation par
rapport au travail.
Un postulat important du modèle DR est que deux processus distincts bien que
reliés, en raison du présumé lien causal entre l'épuisement et le cynisme, sont en
cause dans le développement de l’épuisement professionnel (cf. Schaufeli & Bakker,
2004). Le processus énergétique origine des demandes et est centré principalement
sur la dimension épuisement du syndrome. L’individu qui se sent confronté sans
cesse à des demandes élevées qui exigent de lui un effort supplémentaire constant
voit éventuellement ses ressources énergétiques s’épuiser. Ainsi, des demandes
élevées tendent à réduire les capacités de l’employé sur le plan énergétique et à le
conduire vers un état d'épuisement. Quant à lui, le processus motivationnel origine
des ressources et est centré principalement sur la dimension cynisme du syndrome.
Les ressources augmentent la motivation extrinsèque au travail en facilitant l’atteinte
des objectifs et donc l'accès à des récompenses extrinsèques valorisées. De plus, les
ressources spécifiques rattachées à l’exécution du travail qui satisfont les besoins
fondamentaux d’autonomie, de compétence et d’attachement interpersonnel
favorisent la croissance personnelle et augmentent la motivation intrinsèque au
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
7
travail. Ainsi, des ressources insuffisantes réduisent la motivation au travail et tendent
à engendrer une attitude de désengagement vis-à-vis du travail à accomplir. Cette
attitude de désengagement envers le travail, ou cynisme, est également vue comme
une stratégie d'adaptation défensive centrée sur les émotions visant à mettre un terme
à un état persistant d'épuisement (Cherniss, 1980; Leiter & Maslach, 1988; Shirom,
2003), ce qui expliquerait le présumé lien causal entre l'épuisement et le cynisme.
En bref, le modèle DR propose que l’épuisement professionnel se développe
lorsque l’employé se sent contraint d’atteindre les standards de performance établis
par l’organisation (et parfois sa profession) en dépit de conditions de travail
incapacitantes (demandes élevées) et démotivantes (ressources insuffisantes). La
notion d’équilibre entre la contribution que l’employé apporte à l’organisation et la
contribution que l’organisation apporte à l’employé, en termes de satisfaction de ses
besoins extrinsèques et intrinsèques, est au cœur de la problématique. Le déséquilibre
naît du fait que l’employé donne beaucoup plus à l’organisation qu’il ne reçoit d’elle.
Une exposition prolongée au stress chronique entraîne une liquidation des ressources
énergétiques de l’employé et encourage chez lui le développement d'une attitude
cynique à l'égard de son travail. Éventuellement, une faible réserve énergétique et une
indifférence vis-à-vis du travail affecteront négativement le niveau de sa contribution
à l’organisation (ex : absentéisme, diminution de la performance globale). De plus,
l’épuisement
professionnel
s’accompagne
généralement
de
troubles
psychosomatiques. Des problèmes de santé persistants peuvent même inciter les
travailleurs à quitter leur organisation voire même leur profession.
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
2.2
8
Objectifs de recherche et aperçu des modèles de recherche
2.2.1 Premier article
Dans le premier article intitulé Determinants of burnout: A new taxonomy of
job demands and resources, nous cherchons à clarifier comment les différents
facteurs de stress évoqués dans la littérature sont interreliés et contribuent au
développement de l'épuisement professionnel. Partant du modèle DR et de sa
conceptualisation des environnements de travail, nous proposons une nouvelle
taxonomie de facteurs de stress associés à la vie au travail avec pour objectif
d'approfondir la compréhension de l'ensemble des facteurs qui devraient être
considérés lors de l'élaboration d'un plan de prévention de l'épuisement professionnel
à l'échelle de l'organisation. Notre étude vise à démontrer comment les demandes et
les ressources associées à différentes facettes de l'environnement de travail, qu'ils
s'agissent de facteurs de stress externes situés dans l'environnement ou de facteurs de
stress internes situés à l'intérieur de l'individu mais induits par l'environnement, sont
interreliées et contribuent au développement de l'épuisement professionnel.
Ainsi, nous présumons que les relations entre les facteurs de stress d'un même
type, i.e. soit des demandes ou des ressources, associés à une même facette de
l'environnement sont plus fortes que les relations entre les facteurs de stress d'un
même type associés à des facettes différentes de l'environnement. Les demandes
externes telles que perçues et les demandes internes induites par l'environnement de
travail incluent dans notre modèle de recherche sont 1) liées à la tâche ou au rôle:
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
9
surcharge quantitative et conflit interne dû à une incompatibilité entre les demandes
reliées à la tâche ou au rôle et les demandes reliées à la vie personnelle; 2) liées aux
interactions sociales: hostilité du supérieur immédiat et le sentiment d'être traité
injustement sur le plan interpersonnel par le supérieur immédiat; et 3) liées à
l'organisation: gestion inefficace du changement et pensées liées au départ volontaire.
Les ressources externes telles que perçues et les ressources internes induites par
l'environnement de travail incluent dans notre modèle de recherche sont 1) liées à la
tâche ou au rôle: la clarté de rôle et le degré ressenti d'autonomie au travail; 2) liées
aux interactions sociales: le leadership transformationnel et le sentiment d'être
supporté par le supérieur immédiat; et 3) liées à l'organisation: les pratiques de
gestion mobilisatrices (i.e. développement, partage d'information, reconnaissance non
monétaire) et l'engagement affectif organisationnel. De plus, nous postulons que les
demandes et les ressources internes contribuent directement au développement de
l'épuisement professionnel et jouent un rôle médiateur dans les relations entre les
facteurs de stress externes tels que perçus par les individus et l'épuisement
professionnel.
2.2.2 Deuxième article
Dans le deuxième article intitulé The burning-out employee’s choice between
sickness absenteeism and presenteeism: The boost effect of organizational vision and
excellence values, nous investiguons l'impact des symptômes d'épuisement
professionnel chez les employés qui se présentent au travail sur la performance
globale au travail. Aussi, nous examinons dans quelle mesure une faible performance
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
10
globale au travail, qui dit-on signalerait le présentéisme, est un précurseur de
l'absentéisme pour cause de santé. Enfin, nous examinons le rôle joué par la vision
organisationnelle et deux valeurs organisationnelles liées à l'excellence, la qualité et
l'innovation, dans les processus énergétique et motivationnel menant à l'épuisement
professionnel et à la décision de participer aux activités organisationnelles. Nous
faisons l'hypothèse que la promotion d'une culture d'excellence tend à diminuer la
probabilité des comportements futurs d'absentéisme pour cause de maladie (i.e. plus
grande probabilité de comportements futurs de présentéisme), que l'employé ait
développé des symptômes d'épuisement ou de cynisme.
2.2.3 Troisième article
Finalement, dans le troisième article intitulé Burnout among nursing staff and
intention to leave the profession: Validation of the job demands-resources model,
nous étudions le rôle de l'épuisement professionnel dans les relations entre les
facteurs de stress reliés au travail et à l'environnement social du personnel infirmier et
l'intention de quitter la profession infirmière. Aussi, nous examinons la nature de la
relation entre l'épuisement professionnel et l'intention de quitter la profession
infirmière.
Dans les deux premiers projets de recherche menés auprès de travailleurs
œuvrant au sein d'établissements de santé, l'épuisement professionnel a été mesuré à
l'aide du Maslach Burnout Inventory - General Survey (MBI-GS) (Schaufeli, Leiter,
Maslach, & Jackson, 1996), cet instrument pouvant être utilisé dans n'importe quel
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
11
contexte occupationnel. Pour ce troisième projet de recherche, puisqu'il a été mené
auprès du personnel infirmier exclusivement, nous avons plutôt mesuré l'épuisement
professionnel à l'aide du traditionnel Maslach Burnout Inventory - Human Services
Survey (MBI-HSS) (Maslach & Jackson, 1996), cet instrument ayant été conçu
spécifiquement pour les professionnels qui sont en contact direct avec les
bénéficiaires de services sociaux. Ainsi, nous avons utilisé les deux dimensions clés
de l'épuisement professionnel qui correspondent parallèlement à celles du MBI-GS,
i.e. l'épuisement émotionnel plutôt que l'épuisement et la dépersonnalisation plutôt
que le cynisme.
En regard de notre modèle de recherche, nous postulons que les demandes
favorisent indirectement la dépersonnalisation via l'épuisement émotionnel, et que les
ressources sont principalement associées à la dépersonnalisation et influencent
l'épuisement émotionnel dans une moindre mesure que les demandes. Les demandes
incluent dans notre modèle de recherche sont la surcharge quantitative, le stress lié au
rôle (ambiguïté de rôle et conflit de rôle), l'hostilité des médecins et des patients et
l'interférence du travail sur la vie familiale. Les ressources considérées dans notre
étude sont l'habilitation psychologique (i.e. le sentiment de compétence, l'autorité
décisionnelle, le sens du travail et l'impact du travail), le support du supérieur
immédiat et des collègues et la reconnaissance des médecins et des patients. L'étude
apporte une contribution importante à la compréhension du rôle des facteurs de stress
dans le développement de l’épuisement professionnel chez le personnel infirmier. En
plus de ceux fréquemment étudiés, tels que la surcharge quantitative et le support du
supérieur et des collègues, notre recherche investigue le rôle de facteurs peu
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
12
documenté au niveau empirique : l’habilitation, en tant que ressource psychologique
liée au travail, de même que la reconnaissance et l’hostilité des médecins et des
patients, deux facteurs de stress associés à l’environnement social du personnel
infirmier. Enfin, nous postulons que les deux dimensions clés de l'épuisement
professionnel sont liées indirectement à l'intention de quitter la profession infirmière
à travers les troubles psychosomatiques, associés au processus énergétique, et à
travers l'engagement affectif professionnel, associé au processus motivationnel.
Un dernier commentaire s'impose quant aux rôles différents attribués dans le
premier et le troisième articles à l'engagement affectif et à l'intention de quitter en
regard du processus d'épuisement professionnel. Dans le premier article,
l'engagement et l'intention de quitter (l'expression “thoughts about leaving” est
préférée et utilisée dans cet article1) sont décrits comme des déterminants de
l'épuisement professionnel, plus spécifiquement comme des facteurs de stress internes
induits par l'environnement de travail ou des états psychologiques. À l'inverse,
l'engagement et l'intention de quitter sont traités comme des conséquences de
l'épuisement professionnel dans le troisième article, mais sont également définis en
tant qu'états psychologiques.
1
Tel que noté par Armitage et Conner (2001), dans le cadre des applications de la théorie de l'action
raisonnée de Fishbein et Ajzen (1975), les chercheurs n'ont pas toujours utilisé des instruments qui
mesurent clairement le construit d'intention. Warshaw et Davis (1985) ont relevé différentes façons
selon lesquelles le construit d'intention a été mesuré et différencié les mesures d'intention
comportementale (i.e. “I intend to perform behaviour X”) des mesures d'auto-prédiction (i.e. “How
likely is it that you will perform behavior X?”). Parce que les items utilisés pour mesurer l'intention de
quitter dans le premier et le troisième articles incluent des mesures d'auto-prédiction (e.g., “It is
possible that I leave the hospital within the next year”), nous préférons désigner l'échelle à l'aide de
l'expression “thoughts about leaving”. Nous avons utilisé cette désignation dans le premier article
plus particulièrement car le sens qu'elle suggère nous semblait plus cohérent avec le traitement de
l'intention de quitter l'organisation à titre de facteur de stress favorisant l'épuisement professionnel.
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
13
Si l'engagement et l'intention de quitter sont définis de la même façon dans les
deux articles, nous attribuons de fait des rôles différents à ces variables en regard du
processus d'épuisement professionnel. La différence entre le premier et le troisième
article est l'objet à l'endroit duquel l'individu doit évaluer son état d'esprit.
L'engagement et l'intention de quitter sont des états d'esprit reliés à l'organisation
dans le premier article et à la profession dans le troisième article. Selon nous, les
attitudes défavorables envers l'organisation (i.e. faible engagement et pensées liées au
départ volontaire) sont sujettes à être des réactions à court terme (mais non toujours
directes) à des facteurs situationnels et susceptibles de se manifester au début du
processus d'épuisement professionnel. Les attitudes négatives à l'endroit de
l'organisation sont donc considérées comme des facteurs de stress (internes)
contribuant au développement des symptômes d'épuisement professionnel, soit de
l'épuisement (principalement au niveau émotionnel) ou du cynisme. Lorsque les
symptômes s'intensifient, l'individu devient alors affligé par des problèmes de santé
plus sérieux et persistants qui peuvent l'inciter à considérer la difficile décision de
quitter sa profession. Le processus d'érosion de l'engagement professionnel est aussi
susceptible de progresser dramatiquement à ce stade d'épuisement professionnel.
Morrow (1993) soutient que l'engagement professionnel est une forme relativement
stable d'engagement qui est moins influencée par les facteurs situationnels en
comparaison avec l'engagement affectif organisationnel. Ainsi, une diminution
marquée de l'engagement professionnel est plus susceptible de se produire tard dans
le processus d'épuisement professionnel, alors que l'individu aura été exposé au stress
pendant un certain temps et qu'il lui semblera improbable que l'environnement de
travail ou le contexte occupationnel puisse changer. Ainsi, les attitudes négatives à
Chapitre 2: Cadre théorique, objectifs et aperçu des modèles de recherche
14
l'endroit de la profession (i.e. faible engagement et intention de quitter) sont
considérées comme des conséquences de l'épuisement professionnel.
CHAPITRE 3: MÉTHODOLOGIE
Les trois projets de recherche ont été menés auprès de trois différents
échantillons de répondants. Par ailleurs, ces répondants sont tous des travailleurs
œuvrant au sein d'établissements rattachés au réseau canadien public de la santé. Pour
le premier et le deuxième projets de recherche, nous avons utilisé des échantillons de
convenance composés de travailleurs de différents champs occupationnels. Pour le
troisième projet de recherche, un échantillon aléatoire de personnel infirmier autorisé
a été utilisé. Les données sont corrélationnelles et ont été collectées à l’aide d’un
questionnaire pré-testé et auto-administré, sauf pour le deuxième projet de recherche
pour lequel nous avons utilisé des données informatiques sur les absences passées et
futures compilées par l’établissement de santé. Les données ont été analysées à l'aide
des méthodes d'équations structurelles pour le premier (N= 956) et le troisième
(N=1636) projets de recherche. Pour le deuxième projet de recherche (N=458), nous
avons réalisé des analyses de régression multiple et de régression logistique.
CHAPITRE 4: PREMIER ARTICLE
Determinants of burnout: A new taxonomy of job demands and resources
Geneviève Jourdain and Denis Chênevert2
Abstract
We propose a new taxonomy of job stressors which distinguishes between external
demands and resources originating in the work environment, which has three facets
(i.e. aspects linked to the task/role, to social interactions and to the organization),
from internal demands and resources originating in the person, but induced by the
work environment and associated with its three facets. The research model was tested
on cross-sectional data collected from 956 hospital workers who responded to a selfadministrated questionnaire. The hypothesized relationships between perceived
external and work-induced internal demands associated with the same work
environment facet, and between perceived external and work-induced internal
resources associated with the same work environment facet were all confirmed. In
addition, our presumption that relationships between stressors of the same type, either
demands or resources, associated with the same work environment facet would be
stronger than relationships between stressors of the same type associated with
different work environment facets was generally supported. Regarding the postulated
mediation effects of work-induced internal stressors in the relationships between
2
GJ reviewed the literature, conceived the research model and the questionnaire, analysed and
interpreted the data and wrote the paper. DC supervised the research project and commented on the
questionnaire and the manuscript.
Chapitre 4: Premier article
17
perceived external stressors and the two burnout dimensions, six tests out of eight
confirmed the presence of a mediation effect. Finally, the results suggest that
demands linked to the task or role and resources linked to the organization may be the
most important contributors to the development of burnout among hospital workers.
Chapitre 4: Premier article
4.1
18
Introduction
Organizational stress academics and practitioners have long sought to identify
which subsets of factors - called stressors (Kahn & Byosiere, 1992) - are more likely
to generate strain and eventually burnout in most workers. Many studies have used a
laundry-list approach to predict burnout (Bakker & Demerouti, 2007), lacking an
overarching theoretical framework that could provide a rationale for the choice of
factors to investigate. Recently, Demerouti et al. (2001) proposed a more structured
approach to work environments, in the form of the job demands-resources (JD-R)
model, which emphasizes the role played by stressors (either demands or lack of
resources), linked to different facets of the work environment, in the development of
burnout. We feel that their conceptual approach to work environments could be
enriched by differentiating between perceived objective characteristics of the
environment and internal conditions originating in the individual but induced by the
work environment. The general aim of the present research is to clarify the nature of
the various work-related stressors that have been evoked in the literature and to
examine the connections between them and their role in the burnout process.
Building on the JD-R model, we propose a new taxonomy of job stressors and
demonstrate how demands and insufficient resources associated with different work
environment facets, whether perceived objective characteristics of the work
environment or internal conditions originating in the individual but induced by the
work environment, relate to one another and contribute to burnout. Contrary to most
studies conducted on the JD-R model, our research seeks to understand the particular
Chapitre 4: Premier article
19
influence of specific demands and resources on the two burnout components, rather
than the influence of demands and resources as two higher-order constructs.
4.2
Theoretical Framework and Hypotheses
4.2.1 Determinants of burnout in the job demands-resources model
The JD-R model embraces the dominant theoretical perspective put forth by
Lazarus and Folkman (1984). Burnout is seen as a persistent dysfunctional state that
results from prolonged exposure to chronic stress, that is a situation where a person
feels incessantly confronted with a high level of demands and insufficient resources
linked to work itself and to the context in which the work takes place (Buunk, de
Jonge, Ybema, & de Wolff, 1998; Cooper, Dewe, & O'Driscoll, 2001; Leiter &
Maslach, 2004). One central assumption of the JD-R model (Demerouti et al., 2001)
is that, although each occupation has its own specific risk factors associated with
burnout, these external factors, traditionally called stressors, can be classified into
two general categories, namely job demands and job resources.
Demands refer to those psychological, social, organizational and physical
aspects of the job that require an effort by the employee and are therefore associated
with certain physiological and/or psychological costs (Bakker et al., 2004). Resources
refer to those psychological, social, organizational and physical aspects of the job that
are intended to facilitate the attainment of work goals, to stimulate personal growth or
to reduce job demands (Bakker et al., 2004). The JD-R model emphasizes the role
Chapitre 4: Premier article
20
played by external work-related stressors in the development of burnout, which can
be components of a microsystem (smallest social unit of organized work) or of a
mesosystem (larger complex of smaller work units that comprise an organization).
Carroll and White (1982) argue that an ecological model of burnout also
includes elements of the exosystem (non-work eco-systems that directly impact the
organization and the worker), the macrosystem (larger cultural and world-wide
complex) and the person system (worker). Perhaps because burnout is viewed as a
work-related syndrome, the role of more macro variables, such as exosystem and
macrosystem stressors and of stressors originating in the individual in the
development of burnout, has received less attention. Following the social information
processing theory (Salancik & Pfeffer, 1978), Lee and Ashforth (1991) argue that
macro variables tend to become phenomenologically meaningful and affect the
employee through their relationship with more micro variables associated with the
job. The theory invoked posits that individuals are concerned with defining and
managing their specific work context and will look for cue socially constructed
meanings about that context. Because micro variables associated with the job are
more concrete and salient to the employee, their impact on attitudes and behaviors at
work should be greater than that of macro variables. The lesser attention given to
stressors originating in the individual, which could either be internal negative
conditions or demands (e.g., need level), or lack of internal positive conditions or
resources (e.g., self-esteem), might also be due to the fact that empirical research on
contributing factors has found that environmental variables are more strongly
predictive of burnout than are individual ones (Maslach, 1998). Burnout researchers
Chapitre 4: Premier article
21
usually integrate individual, internal factors that function as personal demands or
resources either as moderator or control variables into their respective model,
assuming these factors are relatively stable or situation independent. Kahn and
Byosiere (1992) argue that stress theorists and researchers have been more interested
in the subset of stimuli that evokes responses of stress in most individuals than in
personal factors that would explain the extent of vulnerability to those stimuli.
Although the JD-R model challenges existing stress models with its
conceptual simplicity and flexibility, it only takes into account direct effects of job
demands and resources on burnout core symptoms without acknowledging the
mediating role of factors and mental processes within the person. The JD-R model
closely resembles the stimulus-response model of human behavior, which is the basis
of behaviorism theory. Behaviorism is based on the proposition that all things which
organisms do - including acting, thinking and feeling - can and should be regarded as
behaviors (Skinner, 1984). However, behaviorists focus solely on observable
behaviors, which are the empirical reality. Without denying the reality of the
individual and internal processes, behaviorists do not examine the role of
psychological constructs considered unobservable and hypothetical. Hence, the
individual is considered as a black box. In this paper, we aim to enrich the JD-R
model conceptual approach to work environments by integrating, in this theoretical
framework, elements of the person, induced by the work environment, which
potentially act as internal stressors mediating the relationships between external
work-related stressors and long-term physical and psychological reactions to chronic
stress, such as burnout symptoms. The paper integrates notions from the JD-R model
Chapitre 4: Premier article
22
with earlier theorizing in the literature (i.e., Hackman, 1977; McGrath, 1976) and
addresses the mediational role played by situationally determined personal demands
and resources in the burnout process.
Our position is based on the premises of the job characteristics model
(Hackman, 1977). This model posits that five core job dimensions create
psychological states that, in turn, produce a number of beneficial personal and work
outcomes. These critical psychological states related to the job (i.e. experienced
meaningfulness, experienced responsibility for outcomes, knowledge of results) are
seen as internal rewards (thus, as valued internal resources) that are obtained when
the worker learns that he personally has performed well on a task that he cares about.
Further, the model posits that the links among the job dimensions, the psychological
states, and the outcomes are moderated by individual growth need strength (which
functions as an internal demand). Hence, the job characteristics model stipulates two
types of internal conditions. First, there are the internal conditions that are distinctive
features of an individual, attributes that people bring with them to the situation, or
“import” into the workplace, such as growth need strength. These types of internal
conditions, which we will term personal conditions (demands or resources), are of
interest if the aim of the research is to identify factors of vulnerability to external
conditions. Secondly, there are the internal conditions which are induced by the work
environment, such as the critical psychological states related to the job (which
function as internal resources). These types of internal conditions, which we will term
work-induced internal conditions (demands or resources), originate in the individual
but are induced by the work environment. Because they are work-related
Chapitre 4: Premier article
23
psychological states and would appear to play a critical role in a number of personal
and work outcomes, these internal conditions have been considered and investigated
by researchers using the JD-R model. However, like organizational stress models, the
JD-R model of burnout does not distinguish external demands (e.g., workload) and
resources (e.g., role clarity) originating in the work environment from internal
demands (e.g., work-family conflict) and resources (e.g., sense of competence)
originating in the person, but induced by the work environment. In the JD-R model,
these different types of demands and resources are considered as intervening at the
same stage in the burnout process, as if all types of demands and resources had a
direct influence on burnout symptoms. Rather, we propose that work-induced internal
demands and resources play a mediating role in the relationships between perceived
external stressors and burnout.
4.2.2 A new taxonomy of demands and resources
4.2.2.1 Determinants of burnout
As shown in Figure 1 (p.25), our revised JD-R model indicates that two types
of external stressors - demands and lack of resources - originate in the work
environment and are linked to three different facets: 1) to the task or the role, 2) to
social interactions, and 3) to the organization (including the physical environment).
These external stressors, as perceived by the individual, become part of his
psychological environment (cf. Lewin, 1951) and induce specific internal demands
and resources originating in the person, or psychological states (i.e. cognitions,
Chapitre 4: Premier article
24
affects and conations). In order to validate our taxonomy, we choose to examine the
connections between work-related stressors that have been frequently evoked in the
burnout literature. We first categorized them according to the definitions given
previously of a job demand and a job resource and classified them either as an
external objective condition of the work environment that can be perceived by any
employee or as an internal psychological state that can only be described by the
employee experiencing it. We then selected pairs of stressors for which some sort of
evidence supporting their relationship existed. The perceived external and related
work-induced internal demands included in our research model are 1) associated
with the task or role: quantitative overload and the internal conflict due to
incompatible tasks or role demands and personal life demands; 2) associated with
social interactions: supervisor hostility and the feeling of being unfairly treated at an
interpersonal level by the supervisor; and 3) associated with the organization:
ineffective change management and thoughts about leaving the organization. The
perceived external and related work-induced internal resources considered in our
study are 1) associated with the task or role: role clarity and a felt degree of
autonomy; 2) associated with social interactions: transformational leadership and the
feeling of being supported by the supervisor; and 3) associated with the organization:
high-involvement management practices (i.e. training, information sharing, nonmonetary recognition) and organizational affective commitment.
Figure 1
Research model
STRESSORS : (LACK OF) RESOURCES
Perceived external resources
(psychological environment)
Work-induced internal resources
(psychological states)
H2a
Task or role
Role clarity
Task or role
Felt degree of autonomy
H5a
H6a
Social interactions
Transformational leadership
H2b
Social interactions
Feeling of being
supported by supervisor
H5b
Cynicism
H6b
Organization
Training practices
Information sharing practices
Non-monetary recognition practices
External
objective
conditions
of the work
environment
H2
(c-e)
H5c
Organization
Affective organizational commitment
H6(c-e)
STRESSORS: DEMANDS
Perceived external demands
(psychological environment)
Task or role
Quantitative overload
H7
Work-induced internal demands
(psychological states)
H1a
Task or role
Internal conflict due to incompatible
tasks or role demands
and personal life demands
H3a
H4a
Social interactions
Supervisor hostility
H1b
Social interactions
Feeling of being unfairly treated at an
interpersonal level by supervisor
H3b
H4b
H3c
Organization
Ineffective change management
H1c
Organizational
Thoughts about leaving
H4c
Exhaustion
Chapitre 4: Premier article
26
The inclusion of thoughts about leaving the organization and lack of affective
commitment as stressors in our framework might seem odd because these
psychological states have usually been treated as consequences of burnout (e.g.,
Burke & Richardsen, 2000; Lee & Ashforth, 1996; Schaufeli & Enzmann, 1998).
Cropanzano, Rupp, and Byrne’s (2003) study provides some evidence supporting the
role of thoughts about leaving the organization as a work-induced internal demand
contributing to the development of a state of exhaustion. They found that the
relationship between emotional exhaustion and intention to leave is partially mediated
by affective organizational commitment. However, because a cross-sectional design
was used, an alternative causal mediational chain is possible. A person who feels
confronted incessantly with a high level of demands and insufficient resources is
likely to develop a weak affective bond toward his organization, a state of mind
which in turn may engender thoughts about leaving. These negative thoughts towards
the organization will require the employee to put in additional effort in order to
perform as expected. If the employee is yet unable or unwilling to leave despite his
desire to do so, then mobilizing personal energies in response to this chronic
“psychological demand” would most likely contribute to heightened feelings of
exhaustion. Health problems engendered by a high level of exhaustion may then lead
the burned-out employee to give more serious consideration to leaving and to ttake
action e.g., looking for a job or receiving training related to a career outside the
present occupational field). For organizational commitment, however, some scholars
view this mind-set (cf. Meyer & Herscovitch, 2001) as an individual resource (e.g.,
Greenglass & Burke, 2002; Hochwarter, Perrewé, Ferris, & Guercio, 1999; Mowday,
Porter, & Steers, 1982) that can increase resistance to stress and burnout. We contend
Chapitre 4: Premier article
27
that these two views of the role of organizational commitment in the burnout process,
as an antecedent and as a consequence, are not contradictory. The lack of external
and internal resources associated with the task or role and with social interactions is
more likely to lead directly to disengagement from work or cynicism (one of the
dimensions of burnout), and indirectly to low affective commitment toward the
organization. In contrast, the lack of external resources associated with the
organization, such as high-involvement management practices, is more likely to
reduce directly organizational commitment and to reduce indirectly engagement to
work, as posited in our model.
4.2.2.2 Interrelations between demands and between resources
Our first group of hypotheses concerns the relationship between perceived
external and work-induced internal demands associated with the same work
environment facet and the relationship between perceived external and work-induced
internal resources associated with the same work environment facet. Our
presumption of the existence of a link between perceived external and work-induced
internal stressors of the same type associated with the same work environment facet
is based on McGrath’s (1976) model of organizational stress. This model posits that,
for each of the six sources of objective stressors (e.g., task, social environment), there
is an equivalent array of subjective or perceived stressors, and for any given
perceived stressor, there is an array of potential responses. In McGrath’s model, the
set of potential responses “nested” within the different perceived sources of stress are
thought to be coping behaviors and not psychological states per se. Nevertheless,
Chapitre 4: Premier article
28
Hackman’s (1977) job characteristics model provides a theoretical basis for the
distinction made between external demands and resources originating in the work
environment that can be assessed by any individual and internal demands and
resources originating in the person, but induced by the work environment. In
Hackman’s model, the five core job dimensions, which would be considered as job
resources linked to the task or role according to the JD-R model, foster the emergence
of critical psychological states related specifically to the task. For example, skill
variety induces experienced meaningfulness of work, which in turn favors high
internal work motivation. While McGrath contends that the six sources of stress in
his model are not totally separate from one another, the relationships between
stressors of the same type associated with the same work environment facet are
thought to be stronger than relationships between stressors of the same type
associated with different work environment facets.
Numerous studies present empirical evidence supporting our first group of
hypotheses. Hypothesis 1 concerns the relationships between perceived external and
work-induced internal demands associated with the same work environment facet.
Beginning with the hypothesized link between demands associated with the task or
role (H1a), many studies report a correlation, ranging approximately from .30 to .60,
between quantitative overload and work interference with home/family conflict (e.g.,
Bakker, Demerouti, & Euwema, 2005; Carlson, Kacmar, & Williams, 2000; Frone,
Yardley, & Markel, 1997; Fu & Shaffer, 2001; Janssen, Peeters, De Jonge, Houkes,
& Tummers, 2004; Voydanoff, 2004). Further, using three waves of data, Demerouti,
Bakker, and Bulters (2004) found that Time 1 work pressure was a determinant of
Chapitre 4: Premier article
29
Time 2 (six-week interval) and Time 3 (three-month interval) work interference with
the home situation. Work overload represents the weight of hours, the sacrifice of
time and the sense of frustration with the inability to complete tasks in the time given.
Long working hours, at the expense of other parts of workers’ lives, and the sense
that there is too much to do in too little time help to create an internal conflict due to
incompatible tasks or role demands and personal life demands.
For the social interactions facet of the work environment, studies have also
linked supervisor hostility to interactional justice (H1b). For example, abusive
supervision has been found to be moderately (-.39) to strongly (-.53) associated with
subordinates’ perception of interactional injustice (cf. Aryee, Chen, Sun, & Debrah,
2007; Tepper, 2000). Interactional justice is fostered when decision makers treat
people with respect and dignity and thoroughly explain the rationale for decisions
(Colquitt, 2001). Workers who believe that their supervisor is abusive with them or
behaves rudely toward them are likely to feel that they are being unfairly treated on
an interpersonal level.
For the organization facet of the work environment, we could not find any
study that has explicitly investigated the impact of perception of ineffective change
management on intention to leave the organization (H1c). Nonetheless, Wanous,
Reichers, and Austin (2000) report a strong link (-.46) between cynicism about
organizational change, which reflects a pessimistic viewpoint about change efforts
being successful, and organizational affective commitment, a robust predictor of
intention to leave the organization (cf. Bentein, Vandenberg, Vandenberghe, &
Stinglhamber, 2005). Organizational changes in themselves can be considered to
Chapitre 4: Premier article
30
constitute a job demand because they require adjustment efforts (cf. Xanthopoulou,
Bakker, Demerouti, & Schaufeli, 2007). Some of the most common implications of
ineffective change management are that productivity declines as people become more
consumed with the change being introduced, employees revert back to the old ways
of doing things and become disinterested in both the current and the future state of
the organization, and changes are scrapped and cancelled due to a lack of support
throughout the organization. These negative outcomes are likely to create a sense of
frustration among workers, because of valuable knowledge and energy wasted upon
efforts to adjust, and thus foster thoughts about leaving the organization.
Hypothesis 2 involves the relationships between perceived external and workinduced internal resources associated with the same work environment facet.
Regarding the presumed link between resources associated to the task or role (H2a),
various studies show that role ambiguity is associated with a felt degree of autonomy
(e.g., Fu & Shaffer, 2001; Schulz, Greenley, & Brown, 1995; Spector, 1986;
Spreitzer, 1996) and that the strength of the correlation could be as high as -.59 (cf.
Fu & Shaffer, 2001). Role ambiguity occurs when an individual is unsure about
others’ expectations of himself. As Spreitzer (1996) argues, if people do not know the
extent of their authority and what is expected of them, they will hesitate to act and
feel they lack self-determination (i.e. sense of choice in initiating and regulating
actions).
For the social interactions facet of the work environment, Doucet, Simard, and
Tremblay (2006) report a very strong correlation of .79 between each component of
transformational leadership and the feeling of being supported by the supervisor, a
Chapitre 4: Premier article
31
finding that supports Hypothesis 2b. Transformational leaders are a role model for
ethical conduct. These leaders sacrifice themselves for the organization and put
subordinates’ interests ahead of their own personal interests. This type of behavior,
which is associated with the charisma component of transformational leadership,
should promote a sense, among subordinates, of being supported by their supervisor
(Doucet et al., 2006). Furthermore, transformational leaders encourage followers to
be innovative by suggesting new angles or ways of doing things. This type of
behavior, which is associated with the intellectual stimulation component of
transformational leadership, is likely to provide subordinates with information that
they can use in coping with work and personal problems, and lead them to feel
supported by their supervisor.
For the organization facet of the work environment, high-involvement
management practices seem to be moderately associated with organizational affective
commitment (H2c-d-e), with correlations ranging from .32 to .49, regardless of which
among training, information sharing or reward and non-monetary recognition
practices are implemented (cf. Meyer & Smith, 2000; Paré & Tremblay, 2007;
Vandenberg, Richardson, & Eastman, 1999). High-involvement management
practices have four attributes, providing 1) the power to act and make decisions about
all aspects of work, 2) information about processes, quality, customer feedback, event
and business results, 3) rewards tied to business results and growth in capability and
contribution, and 4) knowledge of the work, the business and total work system
(Vandenberg et al., 1999). The implementation of such human resource management
practices sends a message that the organization is motivated by the desire to create a
Chapitre 4: Premier article
32
climate of concern and caring, to be fair in dealings with employees and to give them
an opportunity to realize their full potential and be successful. According to Meyer
and Allen (1991), work experiences that satisfy employees’ psychological needs to
feel comfortable, both physically and psychologically, within the organization, and
competent in the work role, are likely to foster affective commitment toward the
organization.
Finally, our presumption of the existence of differences in the strength of the
relationships between perceived external and work-induced internal stressors of the
same type associated with different work environment facets is based exclusively on
our analysis of the pattern of relationships between resources because we were unable
to locate any study that analyzed demands in this way. For example, Fu and Schaffer
(2001) found that role ambiguity is more strongly related to a felt degree of autonomy
(-.59) than to a felt degree of support received from a supervisor (-.33). Doucet et al.
(2006) report a stronger association between each of the two components of
transformational leadership and the feeling of being supported by the supervisor
(.79/.79) than with affective organizational commitment (.32/.30). Paré and Tremblay
(2007) observe that high-involvement management practices relate more strongly to
affective organizational commitment (.32 to .42) than to a felt degree of autonomy
(.25 to .33). On the basis of McGrath’s (1976) model of organizational stress,
Hackman’s (1977) job characteristics model and empirical evidence, we formulate
the following hypotheses regarding the new taxonomy of job demands and resources
proposed:
Chapitre 4: Premier article
33
H1: A perceived external demand (H1a: quantitative overload; H1b:
supervisor hostility; H1c: ineffective change management) is positively
and more strongly related to a work-induced internal demand associated
with the same work environment facet (H1a: internal conflict due to
incompatible demands; H1b unfair treatment by supervisor; H1c:
thoughts about leaving).
H2: A perceived external resource (H2a: role clarity; H2b: transformational
leadership; H2c-d-e: high-involvement practices) is positively and more
strongly related to a work-induced internal resource associated with the
same environment facet (H2a: autonomy; H2b: support from the
supervisor; H2c-d-e: affective organizational commitment).
4.2.3 Interrelated dual processes involved in the development of burnout
Our second group of hypotheses pertains to the role played by each type of
stressors originating either in the work environment or in the person on burnout. We
postulate that work-induced internal demands and resources contribute directly to the
development of burnout and mediate the relationships between perceived external
stressors and burnout. The theorists who developed the JD-R model (cf. Demerouti et
al., 2001) adopted a two-dimensional definition of burnout inspired by Maslach and
Jackson (1986). Exhaustion refers to being emotionally drained and physically or
cognitively exhausted by one’s work. Cynicism (or disengagement from work)
reflects indifference or a distant attitude towards work. The JD-R model assumes that
two distinct yet related processes, because of the presumed causal link between
Chapitre 4: Premier article
34
exhaustion and cynicism, contribute to the development of burnout (cf. Schaufeli &
Bakker, 2004). The energetic process originates from demands and is mainly centered
on exhaustion. In contrast, the motivational process originates from resources and is
mainly centered on cynicism. In addition to the main effects of job demands and
resources, the JD-R model proposes that the interaction between job demands and job
resources is important for the development of job strain and motivation as well
(Bakker & Demerouti, 2007). However, we limited our analyses to the examination
of the main effects of job demands and resources on exhaustion or cynicism because
the aim of the paper is not to predict the level of burnout symptoms but to
demonstrate the independent contribution of each category of work-related stressors
that comprise the proposed taxonomy to the development of burnout symptoms.
4.2.3.1 Energetic process
The influence of demands on exhaustion is supported by the conservation of
resources (COR) theory (Hobfoll & Freedy, 1993). According to the COR theory,
individuals will invest the resources they own in order to protect resources, limit loss
of resources or gain resources. The mobilization of personal energies and other
resources to meet chronic demands induces a “loss spiral” phenomenon that affects
reserves of resources because they tend to co-travel in resource caravans (cf. Hobfoll,
2002). When individuals are exposed to chronic demands, a process of depletion and
wearing down of their energy resources develops over time. Because personal
energies, as secondary resources, determine the access and use of primary resources,
such as social support (cf. Hobfoll & Shirom, 2000), a state of exhaustion of personal
Chapitre 4: Premier article
35
energies at the emotional, physical and cognitive levels should coincide with a state
of impoverishment of other kinds of resources. However, the COR theory also states
that because individuals strive to protect themselves from resource loss, workers are
more sensitive to workplace phenomena that translate into losses for them than to
those that result in gains. Demands threaten or cause a loss of resources that
employees consider important, whereas insufficient resources reduce only the
possibilities of increasing their pool of existing resources. Accordingly, individuals
should show a greater reactivity to high demands than to insufficient resources, and
so demands will therefore be a stronger predictor of exhaustion than resources,
especially internal demands, which are more salient to the person.
In support of this theoretical argument, the meta-analysis by Lee and Ashforth
(1996) demonstrates that demands are positively linked to exhaustion (rc = .21 to
.65), whereas resources, such as various types of support and intrinsic motivational
characteristics at work, are negatively but less strongly correlated (rc = -.10 to -.48)
with this dimension. Further, structural analyses of the JD-R model reveal that
demands are positively related to exhaustion while job resources are associated with
exhaustion to a lesser extent, or not at all (cf. Bakker et al., 2004; Demerouti et al.,
2004). Lastly, evidence suggesting that internal demands may play a mediating role
in the relationships between external demands and exhaustion comes from Bakker et
al. (2005), who show that exhaustion is more strongly related to work interference
with home conflict (.55) than to quantitative overload (.39). Demerouti et al. (2004)
have also found a similar pattern of results in their three-wave study. On the basis of
Chapitre 4: Premier article
36
these theoretical arguments and empirical evidence, we formulate the following
hypotheses:
H3: Work-induced internal demands (H3a: internal conflict due to
incompatible demands; H3b: unfair treatment by supervisor; H3c:
thoughts about leaving) are positively associated with exhaustion.
H4: Work-induced internal demands (H4a: internal conflict due to
incompatible demands; H4b: unfair treatment by supervisor; H4c:
thoughts about leaving) mediate the positive relationships between
external demands (H4a: quantitative overload; H4b: supervisor
hostility; H4c: ineffective change management) and exhaustion.
4.2.3.2 Motivational process
Job resources may play either an extrinsic motivational role, because they are
instrumental in achieving goals, or an intrinsic motivational role, because they fulfill
basic human needs. Two theories, the goal setting theory and the self-determination
theory, thus support the role of resources in this motivational process.
According to the goal setting theory (Locke & Latham, 1990), resources
increase extrinsic motivation at work by facilitating the attainment of objectives.
Even though the work might not be intrinsically motivating, the employee may be
willing to expend the needed amount of personal energy to ensure successful work
performance, given adequate resources, because task performance is compensated
with valued extrinsic rewards. Thus, having insufficient resources to face chronically
Chapitre 4: Premier article
37
imposed demands putatively compromises the attainment of a level of performance,
which would give access to valued extrinsic rewards, therefore creating a feeling of
failure and frustration. Demerouti et al. (2001) assert that disengagement from work
may represent an important self-protection mechanism intended to prevent employee
frustration over difficulty in attaining performance objectives.
In contrast, the self-determination theory (Ryan & Deci, 2004) stipulates that
resources related to the execution of work that satisfy the needs for autonomy,
competence and interpersonal attachment favor personal growth and increase
motivation linked to the accomplishment of tasks. Consequently, the lack of
resources related to these basic human needs may encourage disengagement from an
activity that is not intrinsically satisfying.
In support of the two previous theoretical arguments, Blais, Brière, Lachance,
Riddle, and Vallerand (1993) observed that extrinsic motivation in the form of
identified regulation (work is seen as instrumental, but the employee will perform the
task with a sense of choice) and three types of intrinsic motivation (accomplishment,
stimulation, learning) are negatively linked to depersonalization, the equivalent of
cynicism for workers providing social services. Also, the meta-analysis by Lee and
Ashforth (1996) shows that resources, such as various types of support and intrinsic
motivational characteristics of work, are negatively linked to depersonalization (rc =
-.08, to -.46). Lastly, Wallace and Brinkerhoff (1991) provide evidence that internal
resources may play a mediating role in the relationships between external resources
and cynicism. They find that depersonalization is more strongly associated with
autonomy (-.21) than with role ambiguity (.03). Jackson, Turner and Brief (1986)
Chapitre 4: Premier article
38
provide additional evidence for the organizational facet of the work environment by
showing that affective organizational commitment is a better predictor of
depersonalization (-.46) than contingent rewards (-.29), a compensation practice. On
the basis of these theoretical arguments and empirical evidence, we formulate the
following hypotheses:
H5: Work-induced internal resources (H5a: autonomy; H5b: supervisor
support; H5c: affective organizational commitment) are negatively
associated with cynicism.
H6:
Work-induced internal resources (H6a: autonomy; H6b: supervisor
support; H6c-d-e: affective organizational commitment) mediate the
negative relationships between external resources (H6a: role clarity;
H6b:
transformational
leadership;
H6c-d-e:
high-involvement
practices) and cynicism.
Moreover, a causal link between exhaustion and cynicism has been
incorporated into the JD-R model (cf. Bakker et al., 2004), implying that the two
basic processes proposed by the model are not totally independent. Consistent with
the stress-strain-coping paradigm (Lazarus & Folkman, 1984), cynicism or
disengagement from work can be seen as a defensive coping strategy centered on
emotions intended to end a persistent state of exhaustion (Cherniss, 1980; Leiter &
Maslach, 1988; Shirom, 2003) when resources are exhausted (Cordes & Dougherty,
1993). We, therefore, formulate the following hypothesis:
H7: Exhaustion is positively associated with cynicism.
Chapitre 4: Premier article
4.3
39
Methodology
4.3.1 Participants
The study was carried out among workers of a large hospital in the Canadian
public health care sector. The organization distributed a questionnaire and a return
envelope addressed to the research team to each potential subject. Of the 2,244
employees solicited, 1,031 returned a completed questionnaire, resulting in a
response rate of 46%. For this study, employees in management positions were
excluded. The final sample comprises 956 respondents with the following
characteristics3: 68% are female, 42% are between 46 and 55 years old, 71% have
full-time job status, 78% work on a day shift, and 48% have been employed by the
hospital for over 20 years. The 956 respondents belong to the following personnel
categories: 1) nursing and cardiorespiratory personnel (22%), 2) paratechnical,
auxiliary service and trade personnel (31.5%), 3) office personnel (20%), and 4)
health and social services technicians and professionals (26.5%).
4.3.2 Measures
Unless otherwise specified, employees were asked to express their level of
agreement with each statement on a continuum ranging from Strongly disagree (1) to
Strongly agree (7).
3
Percentages reported are calculated based on valid cases.
Chapitre 4: Premier article
40
Demands. Comprising three items, the quantitative overload scale developed
was based on the instrument designed by Caplan, Cobb, French, Van Harrison, and
Pinneau (1980) (e.g., “I regularly feel overloaded by my work”). The supervisor
hostility scale includes two items taken and adapted from the Inventory of Stressful
Events conceived by Motowidlo, Packard, and Manning (1986) to measure the
frequency of stressful events experienced by nursing staff. A sample item is, “My
supervisor is abusive (verbally, psychologically or physically) toward me”.
Respondents were instructed to indicate the frequency of specific behaviors during
the last twelve months, on a continuum ranging from Never (1) to A few times per
week (7). Perceived ineffective change management was measured using two items
developed for this study, inspired by the work of Collerette and Schneider (1996). A
sample reversed item is “At this point in the implementation, I believe that the new
organizational functioning will be effective”. The three items that form the internal
conflict due to incompatible tasks or role demands and personal life demands scale
were taken from the instrument designed by Kopelman, Greenhaus, and Connolly
(1983) (e.g., “My job takes time that I would like to dedicate to my personal life”).
The feeling of being unfairly treated at an interpersonal level by the supervisor was
measured using three reversed items taken and adapted from the instruments
developed by Colquitt (2001) to capture interpersonal justice (e.g., “My supervisor
treats me with respect and dignity”) and informational justice (e.g., “ My supervisor
is candid in his communications with me”). Lastly, the three items that refer to
thoughts about leaving the organization were developed by Meyer, Allen, and Smith
(1993) and adapted to an agreement-disagreement scale (e.g., “It is possible that I
leave the hospital within the next year”).
Chapitre 4: Premier article
41
Resources. The three items that form the role clarity scale were taken from
the instrument designed by Rizzo, House, and Lirtzman (1970) (e.g., “I know what
my responsibilities are”). Transformational leadership was measured using six items
taken and adapted from the Multifactor Leadership Questionnaire (MLQ Form 5X)
(Avolio, Bass, & Jung, 1999). The scale comprises three items pertaining to charisma
(e.g., “My supervisor is a model for me to follow”) and three pertaining to intellectual
stimulation (e.g., “My supervisor encourages me to see things from a different
angle”). The training practices scale, developed by Simard, Chênevert, and Tremblay
(2000), includes three items (e.g., “In our organization, various professional
development activities are available to employees”). The information sharing
practices (4) scale, designed by Simard et al. (2000), consists of two items related to
top-down communication (e.g., “Employees are regularly informed about major
projects in our organization, such as major investments, new technologies”) and two
items related to bottom-up communication (e.g., “The organization regularly seeks
the employees’ opinions”). Non-monetary recognition practices were measured using
three items generated for this study (e.g., “My organization generally congratulates
employees who surpass expectations”). The three items that form the felt degree of
autonomy scale were taken from the instrument developed by Spreitzer (1995) (e.g.,
“I have significant autonomy in determining how I do my job”). The feeling of being
supported by the supervisor was measured using items taken and adapted from the
instrument developed by Eisenberger, Huntington, Hutchison, and Sowa (1986) to
measure perceived organizational support. The three items selected reflect
instrumental/informational support (e.g., “If I had a problem, I feel I could get help
from my supervisor”). The affective organizational commitment scale comprises three
Chapitre 4: Premier article
42
items taken from Meyer et al.’s (1993) instrument (e.g., “This organization has a
great deal of personal meaning for me”).
Burnout. Exhaustion and cynicism were each measured using three items
selected from the Maslach Burnout Inventory – General Survey (MBI-GS)
(Schaufeli, Leiter, Maslach, & Jackson, 1996) on the basis of their factor loading and
adapted to fit a frequency format response scale. Respondents were asked to indicate
the frequency of symptoms experienced in the last twelve months, on a continuum
ranging from Never (1) to A few times per week (7).
Control variables. The following socio-demographic information was
collected in order to measure the influence of these variables on the endogenous
variables included in our model: gender, age, job status, work schedule,
organizational tenure and occupation. These control variables were selected because
some of these biographical characteristics are the most consistently related to burnout
(i.e. gender, age, tenure) (cf. Schaufeli & Enzmann, 1998) or are likely to be linked to
causal factors such as felt degree of autonomy, affective organizational commitment,
and internal conflict due to incompatible tasks or role demands and personal life
demands (i.e. job status, work schedule, occupation).
4.3.3 Analyses
The model was tested using structural equation methods. Consistent with the
approach suggested by Anderson and Gerbing (1988), we assessed the quality of fit
Chapitre 4: Premier article
43
of the measurement model using confirmatory factor analysis, then estimated a
structural model whereby latent variables were regressed on other latent variables, as
specified by our hypotheses, as well as on the appropriate observed measures. We
also assessed the quality of fit of the next most likely alternative model from a
theoretical perspective to the substantive model of interest. The quality of fit of the
models was evaluated based on several indices. Accordingly, a Root Mean Square
Error of Approximation (RMSEA) below .05 is a sign of a good degree of fit, while
values of up to .08 indicate a reasonable error of approximation in the population
(Byrne, 1998). A Normed Fit Index (NFI) and a Comparative Fit Index (CFI) situated
between .90 and 1 also indicate the presence of a well-fitted model (Bentler, 1992;
Hu & Bentler, 1995). We also report the classic chi-square (χ2) statistic to compare
the quality of fit of nested models. Lastly, the postulated mediation effects were
tested according to Baron and Kenny’s (1986) procedure. The significance of these
effects was examined by applying the Sobel z-test.
4.4
Results
4.4.1 Preliminary analysis
Because results obtained with self-reported measures might be confounded by
common method variance, Harman’s one-factor test was performed (cf. Podsakoff &
Organ, 1986) on the 50 scale items used to assess the 16 variables considered in the
present study. Using an eigenvalue greater than the 1.00 cut-off criterion, 11 factors
emerged from the factor analysis. Although Podsakoff and Organ argue that the
Chapitre 4: Premier article
44
results of such a test cannot be interpreted unequivocally, they do nonetheless provide
some evidence that common method variance does not account for most of the
interrelationships between the variables because the items used to assess the
independent, mediating and dependent variables do not tend to load on a single
general factor.
4.4.2 Measurement model
We conducted a first-order confirmatory factor analysis (CFA) by including
all the 16 variables measured with multiple items. In this 16-factor model, each item
was associated with the factor defined a priori, and factor loadings ranged from .57 to
.94 (p .05), thus meeting the minimum threshold of .50 recommended by Roussel,
Durrieu, Campoy and El Akremi (2002). The level of reliability of internal
consistency of the latent variables (Jöreskog’s rho) varied between .76 and .95, which
satisfies the minimum threshold of .70 recommended by Roussel et al. (2002).
Regarding the quality of fit, the indices suggested a satisfactory level of fit [ 2(1055)
= 3394.64, p .001; RMSEA = .050; NFI = .97; CFI = .98]. Next, because theory
justifies the presence of a meta-construct and the two components of the burnout
syndrome are highly correlated ( = .66), we compared our proposed 16-factor model
with a simpler model, a 15-factor model, in which items related to exhaustion and
cynicism are specified to load on the same factor. This 15-factor model provided a
less satisfactory fit [ 2(1058) = 3694.52, p .001; RMSEA = .053; NFI = .96; CFI =
.97] than the proposed 16-factor model. A comparison of these two models indicated
Chapitre 4: Premier article
45
a significant deterioration in the quality of fit (Δ 2(3) = 299.88, p .001). Also, we
examined the possibility that the items measuring stressors of the same type (i.e.
demand or resource) and belonging to the same set (i.e. external or internal stressor),
represent the same latent factor. As shown in Table 1, the strength of the average
correlations between the latent variables ( ), which is under the minimum threshold
of .60 recommended by Roussel et al. (2002), does not suggest the presence of
second-order factors.
4.4.3 Structural model
Table I (p.46) presents the descriptive statistics and correlations between the
variables of the structural model. As can be seen from this table, significant
correlations exist between all variables for which we have postulated a link. Further,
the significant relationships observed between the exogenous variables and the two
burnout dimensions are the first indication that the postulated mediation effects might
exist.
Table I
Descriptive statistics and correlations
Variables
1. Role clarity
2. Autonomy
3. Leadership
4. Support
5. Training
6. Sharing information
7. Recognition
8. Commitment
9. Quantitative overload
10. Internal conflict
11. Hostility
12. Unfair treatment
13. Change management
14. Thoughts about leaving
15. Exhaustion
16. Cynicism
M
SD
5.66
1.04
5.28
1.34
.41**
4.15
1.47
.40**
.27**
5.05
1.60
.54**
.41**
.75**
3.89
1.46
.29**
.28**
.38**
.37**
3.60
1.36
.32**
.29**
.39**
.36**
.63**
2.97
1.53
.20**
.20**
.32**
.29**
.62**
.54**
4.92
1.60
.37**
.34**
.26**
.35**
.34**
.36**
3.99
1.53
-.09*
.04
.01
-.11**
-.06
-.01
-.03
-.03
3.43
1.60
-.13**
-.16**
-.03
-.13**
-.10**
-.15**
-.10**
-.13**
.54**
1.33
.61
-.29**
-.23**
-.50**
-.53**
-.23**
-.22**
-.19**
-.16**
.14**
.12**
2.83
1.56
-.54**
-.37**
-.80**
-.87**
-.34**
-.34**
-.28**
-.28**
.10**
.14**
.64**
4.08
1.33
-.35**
-.32**
-.35**
-.41**
-.34**
-.44**
-.38**
-.47**
.18**
.23**
.22**
.36**
2.47
1.68
-.29**
-.24**
-.14**
-.27**
-.23**
-.21**
-.17**
-.50**
.15**
.29**
.19**
.22**
.26**
3.08
1.36
-.21**
-.22**
-.11**
-.23**
-.14**
-.16**
-.16**
-.20**
.55**
.78**
.24**
.22**
.30**
.38**
2.70
1.42
-.37**
-.33**
-.25**
-.32**
-.24**
-.24**
-.30**
-.48**
.15**
.36**
.25**
.30**
.39**
.61**
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
.32**
.33**
.44**
.24**
.25**
.17**
.32**
-.04
-.11**
-.22**
-.44**
-.27**
-.22**
-.16**
-.30**
.23**
.36**
.23**
.26**
.17**
.30**
.04
-.14**
-.17**
-.32**
-.26**
-.19**
-.17**
-.26**
.68**
.33**
.38**
.30**
.23**
.03
-.02
-.41**
-.71**
-.30**
-.11**
-.09**
-.20**
.32**
.35**
.27**
.33**
-.08*
-.12**
-.45**
-.81**
-.35**
-.23**
-.21**
-.30**
.59**
.57**
.29**
-.04
-.08*
-.18**
-.30**
-.29**
-.18**
-.11**
-.19**
.56**
.32**
-.01
-.14**
-.19**
-.34**
-.43**
-.19**
-.15**
-.24**
.27**
-.02
-.10**
-.16**
-.26**
-.33**
-.12**
-.13**
-.26**
-.02
-.12**
-.12**
-.25**
-.41**
-.44**
-.19**
-.42**
.43**
.07*
.06
.14**
.11**
.42**
.10**
.11**
.13**
.20**
.21**
.61**
.30**
.53**
.16**
.14**
.18**
.19**
.30**
.18**
.19**
.27**
.18**
.24**
.31**
.28**
.46**
.31**
Note. N =956. Pearson’s correlations (r) are reported above the diagonal. Correlations between latent variables ( ) are reported below the diagonal.
* p .05. ** p .01.
.55**
.66**
Chapitre 4: Premier article
47
During the test of the structural model, some of the relationships between the
control variables and the endogenous variables were significant. However, given that
the inclusion of the control variables induced no change in the value of some
estimated parameters, and not a substantial one for some others (a decrease of .01 or
.02 in the strength of most relationships, the greatest change being .04 for the felt
degree of autonomy-cynicism relationship), and given that all relationships remained
significant without the inclusion of control variables, we decided to exclude these
variables from the structural model for parsimony purposes.
The quality of fit of the structural model to the data appears satisfactory
[ 2(1160) = 6997.99, p .001; RMSEA = .079; NFI = .93; CFI = .94], and the
variance explained for each of the endogenous variables is: 17% for autonomy, 56%
for support, 9% for commitment, 31% for internal conflict due to incompatible
tasks/role and personal life demands, 41% for unfair treatment, 6% for thoughts about
leaving, 59% for exhaustion, and 46% for cynicism. We then assessed the quality of
fit of the next most likely alternative model from a theoretical perspective (i.e. from
the original JD-R model's perspective) to the substantive model of interest. Hence, we
specified a structural model where perceived external stressors, in addition to the
postulated indirect links via the work-induced internal stressors, impact burnout
directly. This alternative model indicated a slightly improvement in the quality of fit
[ 2(1152) = 6933.04, p .001; RMSEA = .079; NFI = .93; CFI = .94] which was
found significant from the perspective of the chi-square (χ2) statistic (Δ 2(8) = 64.95,
p .001). However, with large sample sizes, a significant value for the chi-square
Figure 2
Results of the test of the structural model
STRESSORS : (LACK OF) RESOURCES
Perceived external resources
(psychological environment)
Work-induced internal resources
(psychological states)
.41***
Role clarity
Felt degree of autonomy
-.11***
.75***
Transformational leadership
-.06*
Cynicism
.14***
Training practices
External
objective
conditions
of the work
environment
Feeling of being
supported by supervisor
-.34***
Information sharing practices
.22***
Non-monetary recognition practices
.14***
Affective organizational commitment
STRESSORS : DEMANDS
Perceived external demands
(psychological environment)
.56***
Quantitative overload
Internal conflict due to incompatible
tasks or role demands
and personal life demands
.64***
Feeling of being unfairly treated at an
interpersonal level by supervisor
.25***
Thoughts about leaving
Supervisor hostility
Ineffective change management
.57***
Work-induced internal demands
(psychological states)
.73***
.10***
Exhaustion
.23***
Note. Significant structural relationships between exogenous and endogenous variables (N = 956). Parameters are reported in standardized form.
*p .05. ** p .01. *** p .001.
Chapitre 4: Premier article
49
difference test may be obtained even when there is only a trivial difference between
two nested structural models' explanations of the estimated construct covariances
(Anderson & Gerbing, 1988). Therefore, an indication of goodness of fit from a
practical standpoint is useful in conjunction with formal statistical tests. Hence, on
the basis of the RMSEA, NFI and CFI indices, the comparison of the two models
indicated no practical significant improvement in the quality of fit. Consequently,
from a practical standpoint, we considered that our more parsimonious research
model provides adequate explanation. The research hypotheses are discussed based
on the results obtained for the different structural parameters estimated (see Figure 2,
p.48).
Hypothesis 1 posited that a perceived external demand is positively and more
strongly related with a work-induced internal demand associated with the same work
environment facet. Indeed, Figure 2 shows that quantitative overload (H1a) is
positively linked to internal conflict due to incompatible tasks/role demands and
personal life demands, supervisor hostility (H1b) is positively linked to the feeling of
being unfairly treated at an interpersonal level by supervisor, and ineffective change
management (H1c) is positively linked to thoughts about leaving. The second part of
our hypothesis was tested by performing a series of tests on the structural model. One
external demand at a time, we maintained the postulated path with the internal
demand associated with the same work environment facet and added paths with the
internal demands associated with the two other work environment facets. Table II
(p.50) presents the results of these analyses. For quantitative overload (H1a), the path
with unfair treatment was not significant ( = .03, ns) contrary to the path with
Chapitre 4: Premier article
50
Table II
Relationships between perceived external stressors and work-induced internal
Job resources
Job demands
stressors associated with different work environment facets
Work environment
facets
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Task or role
Social interactions
Organization
Perceived external stressors
Quantitative overload
------Supervisor hostility
------Ineffective change
management
Role clarity
------Transformational leadership
------Training practices
----Sharing information
practices
----Non-monetary recognition
practices
Work-induced internal
stressors
Internal conflict
Unfair treatment
Thoughts about leaving
Internal conflict
Unfair treatment
Thoughts about leaving
Internal conflict
Unfair treatment
Thoughts about leaving
Parameter
= .56***
= .03
= .14***
= .08*
= .64***
= .16***
= .16***
= .25***
= .26***
Felt autonomy
Felt support
Organizational commitment
Felt autonomy
Felt support
Organizational commitment
Felt autonomy
Felt support
Organizational commitment
Felt autonomy
Felt support
Organizational commitment
= .45***
= .38***
= .33***
= .18***
= .75***
= .14***
= .21***
= .16***
= .17***
= .21***
= .13***
= .23***
Felt autonomy
Felt support
Organizational commitment
= .14***
= .08**
= .14***
Note. Parameters are reported in standardized form. *p .05. ** p .01. *** p .001.
thoughts about leaving ( = .14, p .001), which was nonetheless weaker than the
postulated relationship with internal conflict (
= .56, p .001). For supervisor
hostility (H1b), the paths with internal conflict ( = .08, p .05) and thoughts about
leaving ( = .16, p .001) were both significant, although much weaker than the
hypothesized relationship with unfair treatment ( = .64, p .001). For ineffective
change management (H1c), the paths with internal conflict ( = .16, p .001) and
unfair treatment (
= .25, p .001) were both significant, but weaker than or
Chapitre 4: Premier article
51
comparable to the postulated relationship with thoughts about leaving ( = .26,
p .001). Overall, the above results support Hypotheses 1a and 1b and partially
support Hypothesis 1c.
Hypothesis 2 posited that a perceived external resource is positively and more
strongly associated with a work-induced internal resource associated with the same
work environment facet. Again, Figure 2 shows that role clarity (H2a) is positively
linked to a felt degree of autonomy, transformational leadership (H2b) is positively
linked to the feeling of being supported by the supervisor, and training practices
(H2c), information sharing practices (H2d), and non-monetary recognition practices
(H2e) are positively linked to affective organizational commitment. The second part
of our hypothesis was tested by performing a series of tests on the structural model.
One external resource at a time, we maintained the postulated path with the internal
resource associated with the same work environment facet and added paths with the
internal resources associated with the two other work environment facets. Table II
presents the results of these analyses. For role clarity (H2a), the relationships with
support ( = .38, p .001) and commitment ( = .33, p .001) were significant but
weaker than the hypothesized relationship with autonomy ( = .45, p .001). For
transformational leadership (H2b), the relationships with autonomy ( = .18, p .001)
and commitment ( = .14, p .001) were significant but less strong than the postulated
relationship with support (
= .75, p .001). For the three high-involvement
management practices (H2c-d-e), the links with autonomy are either stronger
(training:
recognition:
= .21, p .001) or comparable (sharing information:
= .21, p .001;
= .14, p .001) to the hypothesized relationships with commitment
Chapitre 4: Premier article
(training:
52
= .17, p .001; sharing information:
= .23, p .001; recognition:
p .001). The links with support are either weaker (sharing information:
p .001; recognition:
= .08, p .01) or comparable (training:
= .14, p .001; training:
= .13,
= .16, p .001) to the
postulated relationships with commitment (sharing information:
recognition:
= .14,
= .23, p .001;
= .17, p .001). Overall, the above results
support Hypotheses 2a and 2b and partially support Hypotheses 2c, 2d and 2e.
Hypothesis 3 posited that work-induced internal demands are positively
associated with exhaustion. As Figure 2 illustrates, internal conflict due to
incompatible tasks/role demands and personal life demands (H3a), feeling of being
unfairly treated at an interpersonal level by supervisor (H3b), and thoughts about
leaving (H3c) are positively associated with exhaustion. Thus, Hypotheses 3a, 3b and
3c are confirmed.
Hypothesis 4 posited that work-induced internal demands mediate the positive
relationships between external demands and exhaustion. Mediational analyses
revealed that internal conflict (H4a) partially mediates the relationship between
quantitative overload and exhaustion because the direct relationship ( = .46, t =
11.39, p .001) became significantly weaker (z = 8.42, p .001) after the inclusion of
the mediator ( = .17, t = 4.43, p .001). However, unfair treatment (H4b) does not
mediate the relationship between supervisor hostility and exhaustion. The inclusion
of the mediator had no significant influence ( = .13, t = 3.20, p .001; z = 0.67, ns) on
the strength of the direct relationship ( = .14, t = 4.66, p .001). Finally, thoughts
about leaving (H4c) partially mediates the relationship between ineffective change
Chapitre 4: Premier article
53
management and exhaustion because the direct relationship ( = .13, t = 3.57, p .001)
became significantly weaker (z = 4.44, p .001) after the inclusion of the mediator (
= .09, t = 2.95, p .01). Overall, the above results partially support Hypotheses 4a and
4c and refute Hypothesis 4b.
Hypothesis 5 posited that work-induced internal resources are negatively
associated with cynicism. As illustrated in Figure 2, felt degree of autonomy (H5a),
feeling of being supported by the supervisor (H5b), and affective organizational
commitment (H5c) are negatively associated with cynicism. Thus, Hypotheses 5a, 5b
and 5c are confirmed.
Hypothesis 6 posited that work-induced internal resources mediate the
negative relationships between external resources and cynicism. Mediational analyses
revealed that autonomy (H6a) partially mediates the relationship between role clarity
and cynicism because the direct relationship ( = -.14, t = -4.02, p .001) became
significantly weaker (z = 1.64, p=.05) after the inclusion of the mediator ( = -.13, t =
-3.29, p .001), although this mediation effect is small. However, support (H6b) does
not mediate the relationship between transformational leadership and cynicism. The
inclusion of the mediator had no significant influence ( = -.11, t = -2.17, p .05; z =
0.33, ns) on the strength of the direct relationship ( = -.09, t = -2.99, p .01). Next,
affective organizational commitment (H6c) fully mediates the relationship between
training practices and cynicism because the previously significant direct relationship
( = -.08, t = -2.46, p .05) became non-significant after the mediator was added to the
model ( = -.01, t = -0.30, ns; z = 3.46, p .001). Likewise, affective organizational
Chapitre 4: Premier article
54
commitment (H6d) fully mediates the relationship between information sharing
practices and cynicism because the previously significant direct relationship ( = -.07,
t = -2.20, p .05) became non-significant after the mediator was added to the model (
= .01, t = .36, ns; z = 5.18, p .001). Finally, affective organizational commitment
(H6e) partially mediates the relationship between non-monetary recognition practices
and cynicism because the direct relationship ( = -.17, t = -5.37, p .001) became
significantly weaker (z = 3.61, p .0.001) after the inclusion of the mediator ( = -.11,
t = -3.44, p .001). Overall, the above results support Hypotheses 6c and 6d, partially
support Hypotheses 6a and 6e, and refute Hypothesis 6b.
Lastly, as shown in Figure 2, the results confirm the existence of a positive
relationship between exhaustion and cynicism, which supports Hypothesis 7.
4.5
Discussion
This study attempted to validate a new taxonomy of job demands and
resources that rests on the conceptual approach to work environments proposed by
Demerouti et al. (2001), who developed the job demands-resources (JD-R) model of
burnout. The general aim of the present research was to clarify the nature of the
various work-related stressors that have been evoked in the literature and to examine
the connections between them and their role in the burnout process. Unlike most
studies conducted on the JD-R model, our research has examined the unique
influence of specific demands and resources on the two burnout components rather
than the influence of demands and resources as two higher-order constructs. With the
Chapitre 4: Premier article
55
taxonomy proposed, practitioners can gain a greater understanding of the network of
variables that should be considered when planning organizational-level interventions
to prevent burnout.
4.5.1 Interpretation of results
Overall, the results obtained tend to support the new taxonomy of job
demands and resources proposed on the basis of McGrath’s (1976) model of
organizational stress, Hackman’s (1977) job characteristics model and piecemeal
empirical evidence on how stressors of the same type relate to one another and to the
two burnout dimensions. The relationships between perceived external and workinduced internal demands associated with the same work environment facet, and the
relationships between perceived external and work-induced internal resources
associated with the same work environment facet were all significant and in the
expected direction. In addition, our presumption that relationships between stressors
of the same type associated with the same work environment facet would be stronger
than relationships between stressors of the same type associated with different work
environment facets was generally supported. Of the 24 relationships examined in this
study, only 5 were comparable in strength or stronger (one case) than what we had
expected. One of these relationships pertained to job demands, more specifically the
link between ineffective change management and unfair treatment, which was
comparable to the postulated relationship with thoughts about leaving. The other
three relationships concerned job resources, more precisely high-involvement
Chapitre 4: Premier article
56
management practices. Hence, training practices were found to be more or as strongly
linked to a felt degree of autonomy and to the feeling of being supported by the
supervisor than to affective organizational commitment. While for information
sharing and non-monetary recognition practices, the links with a felt degree of
autonomy were found to be comparable to the hypothesized relationships with
affective organizational commitment. These results contradict previous findings that
high-involvement management practices relate more strongly to affective
organizational commitment than to a felt degree of autonomy (cf. Paré & Tremblay,
2007). It is interesting to note that the five relationships that were comparable in
strength or stronger than what we had expected concern perceived external resources
and demands associated with the organization. This finding suggests that external
stressors which are linked to the organization, a more macro facet of the work
environment, can have a considerable influence on internal stressors that are
associated with more micro facets of the work environment, such as the task or role
and the social interactions at the work unit level. Lee and Ashforth (1991) have
indeed suggested that the effect of organizational structure and processes on burnout
is mediated by job-related stressors, such as experienced role pressures (e.g., role
ambiguity and role conflict). Following the social information processing theory
(Salancik & Pfeffer, 1978), hey argue that macro variables tend to become
phenomenologically meaningful and affect the employee through their relationship
with micro variables associated with the job. In contrast, our study provides no
evidence that external stressors which are linked to more micro facets of the work
environment can have a substantial influence on internal stressors that are related to a
larger ecosystem, such as an organization. This pattern of results makes some sense.
Chapitre 4: Premier article
57
For example, human resource management practices are implemented with the aim,
among others, of optimizing employees’ performance in their work role and of
favoring high quality exchange relationships between organizational members.
However, employees may perceive some job demands associated with the task/role or
with proximate social interactions but will not necessarily have thoughts about
leaving the organization, unless they have been exposed for some time to multiple job
demands that are elements of these two more micro facets of the work environment.
Nevertheless, additional studies are needed to confirm this pattern of results.
The results also provide support for the dual processes involved in the
development of burnout. The positive relationships postulated between work-induced
internal demands and exhaustion were all significant. Of the three internal demands
considered in this study, the demand associated with the task or role, i.e. internal
conflict between incompatible tasks/role demands and personal life demands, had the
strongest relationship with exhaustion. Further, the negative relationships
hypothesized between work-induced internal resources and cynicism were all
significant. Of the three internal resources considered in this study, the resource
associated with the organization, i.e. affective organizational commitment, had the
strongest relationship with cynicism. The results suggest that demands associated
with the task or role and resources associated with the organization may be the most
important contributors to the development of burnout among hospital workers.
However, because we validated our taxonomy with only one stressor example
associated with each category and because the study was carried out among workers
employed by one single organization, more studies are needed to confirm these
Chapitre 4: Premier article
58
findings. Finally, exhaustion was found to be positively and significantly associated
with cynicism. Overall, the results confirmed the JD-R model postulate that two
distinct yet related processes - energetic and motivational - contribute to the
development of burnout.
Regarding the postulated mediation effects of work-induced internal stressors
in the relationships between perceived external stressors and the two burnout
dimensions, six of the eight mediational tests conducted confirmed the presence of a
mediation effect. Two of these tests, pertaining to affective organizational
commitment, revealed full mediation effects. Recently, the role played by personal
resources in the motivational process of the JD-R model has been underscored by the
work of Xanthopoulou et al. (2007). Focusing on the improvement of employee wellbeing rather than on negative outcomes variables such as burnout and ill health, they
expanded the JD-R model by showing that personal resources mediate the
relationship between job resources and work engagement. However, their study
examines the influence of job resources (e.g., supervisory coaching, professional
development) treated as a higher-order construct on three personal resources treated
as a unitary resiliency construct. Furthermore, two of these personal resources are
general (i.e. self-efficacy and optimism) rather than situation- and facet-specific.
Likewise, we acknowledge earlier attempt at theorizing the role of personal demands
in the stress process (cf. Siegrist, 1996). However, in regard to the energetic process
of the JD-R model, our study is the first to our knowledge to examine how job
demands associated with different work environment facets influence exhaustion
through work-induced personal demands.
Chapitre 4: Premier article
59
Two postulated mediation effects were not confirmed. We attribute the
absence of mediation effects observed for supervisor support and supervisor unfair
treatment, operationalized as psychological states, to their strong relationships with
the independent variables, namely transformational leadership and supervisor
hostility, respectively. When the path between the independent and the mediating
variables is large, there is less unique variance in the mediating variable to explain
the dependent variable. The strength of these relationships is not surprising given the
fact that the exogenous variables were not external objective conditions of the work
environment, but rather, reflected the psychological environment of the individual,
i.e. perceptions. However, we must underline that the collinearity diagnostics did not
reveal a problematic variance inflation factor (VIF) and condition index for either of
the two multiple regression analyses conducted.
4.5.2 Study limitations
This study has some limitations that should be noted. First, the sample,
although reasonably large, does not allow strong inferences because it represents one
single organization, and we could not examine non-response bias. Secondly, we used
cross-sectional data to examine presumed causal relationships and to verify
mediational hypotheses. Nevertheless, we provided readers with theoretical
arguments and some empirical evidence supporting the temporal order implied by our
mediation model, and we also tested an alternative model. Third, the use of
abbreviated scales and less than three items for some constructs might have affected
our results although the latent variables showed a generally satisfactory level of
Chapitre 4: Premier article
60
reliability of internal consistency. However, the fact that the relationships between
the different variables, based on a priori set hypotheses, were generally confirmed
strengthen our confidence in regard to construct validity. Fourth, it may be argued
that our results obtained with self-reported measures from a single source might be
confounded by common method variance. Indeed, some of the variables included in
this study - mainly pertaining to the supervisor - were highly correlated. However,
these variables may have been highly correlated because of valid functional
relationships among them. While we justify the use of self-reported measures in our
study by the phenomenological definition of stress proposed by Lazarus and Folkman
(1984) and the psychological nature of many constructs included in our research
model, respondents were asked to express their judgment on relatively concrete
attitudes, behaviors or facts, thus leaving little room for inference (cf. Doty & Glick,
1998). Also, we provided some evidence, based on the results of Harman’s one-factor
test, that common method variance might not pose a problem in the present study. As
a potential cause of common method variance, the omission of negative and positive
affectivity in our research model may represent a limitation, because this personality
trait purportedly explains the magnitude of relations observed between the perception
of stress factors and self-reported symptoms (Burke, Brief, & George, 1993; Watson,
Pennebaker, & Folger, 1987). As this trait is a potential cause of the variables
investigated, these studies recommend that its effect be isolated. However, a number
of more recent studies have demonstrated that this affective disposition does not
significantly bias relations between self-reported measures of stress factors and
associated symptoms, to the extent that they are invalidated (e.g., De Jonge,
Chapitre 4: Premier article
61
Dormann, Janssen, Dollard, Landeweerd, & Nijhuis, 2001; Schonfeld, 1996;
Williams, & Anderson, 1994; Williams, Gavin, & Williams, 1996).
4.5.3 Practical implications
The taxonomy proposed and the results obtained from this study suggest that
practitioners should focus their efforts to prevent burnout on factors that are primarily
responsible for the troubled person-environment relationship (stress): perceived
external resources and demands associated with the task or role, with social
interactions and with the organization. According to Lazarus and Launier (1978), two
strategies can serve to reduce or eliminate stress or make it more tolerable. Efforts
could either be directed at altering a stressful person-environment relationship
(problem-focused strategy) or at controlling the emotional reaction arising from that
relationship (emotion-focused strategy). Organizations have the power to implement
actions aimed at changing the task and the work environment (problem-focused
strategy), and to communicate information to employees about available resources
that might not be perceived. This information can then be used to re-appraise the task
and the work environment and help them feel better (emotion-focused strategy) by
making the transaction seem more under control. These two strategies reflect efforts
directed at the environment, how it is objectively, and how it is perceived. Thus, a
comprehensive organizational strategy to prevent burnout would include actions,
especially implementation of human resource management practices whose influence
is more systemic, aimed at increasing job resources to meet job demands associated
Chapitre 4: Premier article
62
with the task or role, with social interactions and with the organization, along with
actions aimed at communicating information about available resources.
4.5.4 Research avenues
Our study attempted to demonstrate how job stressors of the same type
originating in the work environment or in the person are interrelated. Another aspect
of the JD-R model that needs to be explored in greater depth is the role of job
resources in the development of burnout. In addition to their motivational role, one of
the presumed functions of job resources is to reduce demands (cf. Bakker et al.,
2004). Therefore, one worthy research avenue would be to examine the interrelations
between perceived external job resources and perceived external job demands
associated with the same work environment facet. Research has demonstrated that job
resources have typically been more weakly related to burnout than job demands
(Hakanen, Schaufeli, & Ahola, 2008). It is possible that job resources whose effect
has been found to be weak or non-significant may indirectly affect the two core
dimensions of burnout via their impact on job demands. Another interesting question
concerns the dynamics between resources. In the conservation of resources (COR)
theory (Hobfoll & Shirom, 2000), social support is viewed as the major vehicle
through which individuals’ resources are widened beyond the limited domain of
resources that are contained in the self. Social support is a meta-construct (Hobfoll,
2002). It includes the actual receipt of acts that have a supportive intent, which could
result in access to valued resources, such as objects, conditions, personal
characteristics and energies. It also includes the sense of close ties that we have with
Chapitre 4: Premier article
63
others that can provide aid, which would be viewed as a condition resource.
Accordingly, the question arises whether resources associated with social interactions
play a critical role in the prevention of burnout. Resources associated with this facet
of the work environment may give access to resources associated with the task/role
and the organization, which would be more proximal determinants of cynicism. Our
results suggest this might be the case because resources associated with social
interactions tend to be, on average, less strongly related to cynicism than resources
associated with the two other work environment facets. If individuals lacking
resources associated with social interactions may also lack or be unable to obtain
other kinds of resources, more research on situational and personal factors that
determine socially supportive acts would be needed.
CHAPITRE 5: DEUXIÈME ARTICLE
The burning-out employee’s choice between sickness absenteeism and presenteeism:
The boost effect of organizational vision and excellence values
Geneviève Jourdain and Denis Chênevert4
Abstract
One aim of the paper is to investigate the relationships between burnout
symptoms among employees who choose to bring themselves to work despite their
ill-health, an act known as presenteeism, overall job performance and future
presenteeism/absenteeism. Also, we examine the moderating role of organizational
vision and excellence values in the relationships between burnout symptoms among
working employees and future presenteeism/absenteeism. The hypotheses were tested
on data collected from 428 workers of a health and social services center who
consented to their absences file’s being matched to the questionnaire responses. Our
study provides strong support for the proposition that low job performance said to
signal presenteeism is a precursor of sickness absenteeism. Also, the results reveal
that the articulation of an organizational vision and the promotion of quality and
innovation, as perceived by burning-out employees, prompt them to choose
presenteeism over sickness absenteeism in the following year.
4
GJ reviewed the literature, conceived the research model and the questionnaire, analysed and
interpreted the data and wrote the paper. DC supervised the research project and commented on the
questionnaire and the manuscript.
Chapitre 5: Deuxième article
5.1
65
Introduction
Health may be a person’s most valued possession: health problems curtail or
prohibit the basis activities of life, including work (Schultz & Edington, 2007). As a
chronic health problem, burnout develops gradually and results from prolonged
exposure to chronic stress (Buunk et al., 1998; Cooper et al., 2001; Leiter & Maslach,
2004). Chronic stress is thought to affect a person’s capacity and motivation to work
(cf. Schaufeli & Bakker, 2004).
In the process of burning-out, the employee will, at some point, have to
choose whether or not to participate in organizational activities. In the health care
sector, deciding to show up at work despite health problems that adversely affect
performance could have serious implications for quality of care and patient safety, as
several studies on burnout have shown (e.g., Aiken, Clarke, & Sloane, 2002;
Argentero, Dell'Olivo, & Ferretti, 2008; Garman, Corrigan, & Morris, 2002;
Halbesleben,
Shanafelt,
Wakefield,
Wakefield,
&
Cooper,
2008;
Bradley, Wipf, & Back, 2002; Shirom, Nirel, & Vinokur, 2006).
Presenteeism, defined as the act of bringing oneself to work although sick (Aronsson
& Gustafsson, 2005; Kivimäki et al., 2005), also entails risks of aggravating the
health of workers. Indeed, working while ill has been shown to increase the risk of
serious coronary events (cf. Kivimäki et al., 2005).
The current study was conducted in the health care sector, where presenteeism
is known to be prevalent (cf. Aronsson, Gustafsson, & Dallner, 2000). A question that
Chapitre 5: Deuxième article
66
has begun to garner more interest during the last decade concerns the determinants of
presenteeism. While ill health can be treated as a common cause of presenteeism and
sickness absenteeism (Johansson & Lundberg, 2004), studies have shown that many
work-related, social and personal factors influence a person’s choice between these
two alternative types of illness behaviors, such as a lack of replacement personnel,
team designs, draconian absence control systems and lack of paid sick leave, job
insecurity and personal financial insecurity (Johns, 2008). However, to our
knowledge the implications of perceived organizational vision and values on the
burning-out employee’s decision to participate in organizational activities have not
been investigated.
This paper makes a significant departure from the literature by examining the
influence of macro-organizational factors characterizing a culture of excellence,
namely vision and values (cf. Kouzes & Posner, 1987; Peters & Waterman, 1982), on
the decision to participate in organizational activities. If superordinate goals and
values are employed to provide members with direction and a sense of purpose and
can be a powerful way to motivate, instill pride of affiliation, and coordinate
collective action (O’Reilly and Chatman, 1996), such a culture has been criticized for
making the achievement of firm goals the only outcome that matters and
compromising employees’ health (e.g., Aubert & De Gaulejac, 1991; Godard &
Delaney, 2000). In their best-selling 1980 book, Burnout: The High Cost of High
Achievement, Freudenberger and Richelson describe burnout as “the super-achiever
sickness.” Burned-out individuals are portrayed as goal-oriented, determined
idealists, over self-confident and strongly dedicated and committed. With health care
Chapitre 5: Deuxième article
67
establishments around the world facing a shortage of health care staff, an interesting
question is whether employees who are gradually experiencing burnout symptoms
(hereafter referred to as the “burning-out employees”) will decide to attend work if
their organization strongly promotes a culture of excellence by communicating its
vision and emphasizing quality and innovation as organizational values?
The present study has two aims. The first research objective is to investigate
the role played by organizational vision and two organizational values representing
excellence, quality and innovation (cf. Peters & Waterman, 1982), in the energetic
and motivational processes that lead to burnout and the decision to participate in
organizational activities. Specifically, we intend to examine how the articulation of
an organizational vision and the promotion of excellence values prompt burning-out
employees to choose presenteeism, i.e. to go to work despite health problems and
reduced job performance, over sickness absenteeism.
Second, this study addresses a question formulated by Aronsson et al. (2000):
“is there a risk that the sickness presenteeism of today will become the sickness
absenteeism of tomorrow?” (p.509). Presenteeism research has been mostly dedicated
to the investigation of individual productivity loss, or in-role performance problems,
owing to reduced capacity associated with the act of going to work despite health
problems (cf. Mattke, Balakrishnan, Bergamo, & Newberry, 2007). We adopt a
broader perspective by examining how a low level of overall job performance due to
the gradual intensification of burnout symptoms among employees who choose to
bring themselves to work despite their ill-health, an act known as presenteeism,
increases the probability of future sickness absenteeism. Specifically, we seek to
Chapitre 5: Deuxième article
68
understand how working at a reduced capacity and reduced motivation, manifested by
in-role performance problems and low extra-role performance, prompt burning-out
employees to choose sickness absenteeism over presenteeism.
5.2
Theoretical Framework and Hypotheses
As Figure 1 (p.69) illustrates, we postulate that burnout, through the energetic
and motivational processes that contribute respectively to the development of its two
core dimensions - exhaustion and cynicism - affects a person’s capacity (exhaustion)
and motivation (cynicism) to work. As shown in the top part of Figure 1, we posit that
when employees afflicted with burnout symptoms choose to bring themselves to
work despite their ill-health, an act known as presenteeism, they will encounter inrole performance problems if they experience feelings of exhaustion, and show low
extra-role performance if they experience feelings of cynicism. In turn, we propose
that low job performance said to signal presenteeism is a precursor of sickness
absenteeism. Furthermore, as shown in the bottom part of Figure 1, we expect that the
promotion of a culture of excellence, which entails communicating an organizational
vision and emphasizing quality and innovation as organizational values, will lower
the probability of future sickness absenteeism (i.e. higher probability of future
presenteeism), whether the employee is experiencing gradual exhaustion or cynicism.
Chapitre 5: Deuxième article
69
Figure 1
Research model
Choice : Presenteeism
Motivational process
Motivation to work
Cynicism
Consequences of presenteeism
H2
H3
Energetic process
Capacity to work
Exhaustion
H4
H1
Low
extra-role
performance
In-role
performance
problems
H8
H7
Future
absenteeism
Future
absenteeism
Continuum of withdrawal behaviors
Choice : Presenteeism
Motivational process
Motivation to work
Cynicism
Organizational vision
and excellence values
H5
H10
Future
absenteeism
Organizational vision
and excellence values
Energetic process
Capacity to work
Exhaustion
H6
H9
Future
absenteeism
5.2.1 The job demands-resources model of burnout
The job demands-resources (JD-R) model of burnout (Demerouti et al., 2001)
embraces the dominant theoretical perspective put forth by Lazarus and Folkman
(1984). Burnout is seen as a persistent dysfunctional state that results from prolonged
exposure to chronic stress, that is a situation where a person feels incessantly
confronted with a high level of demands and insufficient resources linked to work
itself and to the context in which the work takes place (Buunk et al., 1998; Cooper et
al., 2001; Leiter & Maslach, 2004). Demands refer to those psychological, social,
Chapitre 5: Deuxième article
70
organizational and physical aspects of the job that require an effort by the employee
and are therefore associated with certain physiological and/or psychological costs
(Bakker et al., 2004). Resources refer to those psychological, social, organizational
and physical aspects of the job that are intended to facilitate the attainment of work
goals, to stimulate personal growth or to reduce job demands (Bakker et al., 2004).
The theorists who developed the JD-R model (cf. Demerouti et al., 2001)
adopted a two-dimensional definition of burnout inspired by Maslach and Jackson
(1986). Exhaustion refers to being emotionally drained and physically or cognitively
exhausted by one’s work. Cynicism (or disengagement from work) reflects
indifference or a distant attitude towards work. The JD-R model assumes that two
distinct yet related processes, because of the presumed causal link between
exhaustion and cynicism, contribute to the development of burnout (cf. Schaufeli &
Bakker, 2004). The energetic process originates from demands and is mainly centered
on exhaustion. In contrast, the motivational process originates from resources and is
mainly centered on cynicism. Chronic stress, through the energetic and motivational
processes that contribute respectively to the gradual intensification of burnout two
core symptoms, affects a person’s capacity (exhaustion) and motivation (cynicism) to
work.
Chapitre 5: Deuxième article
71
5.2.2 Presenteeism as a precursor of sickness absenteeism
5.2.2.1 The concept of presenteeism
Although they may be experiencing feelings of exhaustion or cynicism, some
workers may decide to participate in organizational activities by attending work,
albeit with reduced capacity and motivation. Presenteeism is defined as the act of
bringing oneself to work although sick (Aronsson & Gustafsson, 2005; Kivimäki et
al., 2005). There is much confusion in the literature about this construct because some
researchers (e.g., Schultz & Edington, 2007; Mattke et al., 2007) portray
presenteeism as the individual productivity loss associated with the behavior of
attending work with health problems. After noting two complementary definitions in
the literature (cf. Johns, 2008), Johns (2009) adopts a convergent perspective and
defines presenteeism as “showing up to work when one is ill and the decrement in
productivity that fallows from this practice” (p.7). In our view, the behavior of
attending work with health problems, i.e. presenteeism, should not be confounded
with the potential consequence of such a behavior in terms of decreased job
performance.
Hence, from our perspective and in line with the definition given above, the
concept of presenteeism includes two inclusive conditions: 1) having health
problems, and 2) choosing to go to work (cf. Dew, Keefe &, Small, 2005).
Conventionally, researchers have measured presenteeism with a single self-reported
item, asking whether during the last 12 months respondents had gone to work
Chapitre 5: Deuxième article
72
although, for health reasons, they should have taken sick leave (e.g., Aronsson et al.,
2000; Elstad & Vabo, 2008). In our study, it should be noted that burnout is a chronic
illness (Schaufeli & Enzmann, 1998) and respondents were asked to report on their
level of job performance. Therefore, we can infer that respondents who reported some
degree of burnout chose to go to work although sick (i.e. presenteeism), otherwise
they could not have assessed their job performance. Thus, like many studies
conducted on the topic of presenteeism, we examine the impact of a specific chronic
health condition, i.e. burnout, when employees attend work (i.e. when present at work
but ill) on job performance (cf. Schultz & Edington, 2007), and how low job
performance said to signal presenteeism is a precursor of sickness absenteeism.
5.2.2.2 Burnout and job performance
Burning-out employees who decide to participate in organizational activities
despite their health problems, whether they are experiencing feelings of exhaustion or
cynicism, face a decision regarding the type of job performance behaviors in which
they will choose to invest their energy resources and the level of performance they
will be able to and will want to attain. At the most general level, job performance
refers to all the behaviors employees engage in at work that contribute to
organizational goals (Jex, 1998). The various behavioral dimensions that comprise
the domain of job performance (cf. Jex, 1998) can easily be classified into two
general categories: in-role and extra-role behaviors. In-role performance behaviors
are those required by one’s formal job description that are needed to perform core
tasks associated with the job (Jex, 1998). Extra-role performance behaviors, also
Chapitre 5: Deuxième article
73
called organizational citizenship behaviors (OCBs), are discretionary behaviors on
the part of the employee that directly benefit the organization (OCB-O) or
immediately benefit specific individuals (OCB-I) and thus indirectly promote the
effective functioning of the organization (Williams & Anderson, 1991). In the present
study, we examine relationships between the two burnout dimensions and in-role and
extra-role performance behaviors in the form of OCB-O. As mentioned, chronic
stress affects a person’s capacity and motivation to work and these two factors are
known to be among job performance determinants (cf. Jex, 1998). We chose to focus
on OCB-O rather than OCB-I because burnout results from prolonged exposure to a
high level of demands and insufficient resources that are linked to work itself and to
the context in which the work takes place. According to the JD-R model, demands
and resources are associated with different facets of the work environment, namely
the task or the role, social interactions, and the organization (including the physical
environment). Hence, employees’ response to burnout symptoms is more likely to be
directed toward the organization rather than a specific individual.
Employees experiencing gradual exhaustion arising from high and chronic job
demands, who have decided to participate in organizational activities despite their
health condition, will most likely choose to invest their limited energy resources in
the sphere of in-role performance because in-role performance behaviors are
monitored and reinforced by organizational rewards and sanctions systems. However,
they will feel incapacitated because their energy resources are diminished, and will
consequently experience in-role performance problems (Bakker et al., 2004). This
assertion is supported by several studies that show that exhaustion is negatively
Chapitre 5: Deuxième article
74
associated with in-role performance (cf. Taris, 2006). The compensatory control
model of performance regulation (Hockey, 1997) provides a theoretical explanation
for the postulated relationship between exhaustion and in-role performance problems.
According to this model, primary task performance may be protected under stress by
the recruitment of further resources, but only at the expense of increased subjective
effort, and physiological and psychological costs. Alternatively, stability can be
achieved by reducing performance goals, without further costs. Therefore, confronted
with high demands, individuals may first try to maintain their level of in-role
performance by mobilizing additional resources at the expense of an increase in
energy costs, psychophysiological activation and fatigue. Although this strain coping
mode can be adaptive over the short term, the chronic use of this type of response
gradually leads to exhaustion because there is little opportunity for recovery from the
fatigue associated with such a coping mode. Hence, when high demands are chronic
and energy is low, an alternative and more appropriate response is to adopt the less
demanding passive coping mode involving downwards adjustment of performance
targets. Lower performance standards may be set, for example, by reducing required
levels of accuracy or speed or by paying less attention to subsidiary activities. By
deploying effort well within reserve limits, this strategy does not incur further energy
costs. However, the reserve effort budget may be set quite low when overall capacity
has been compromised by chronic stress. A small reserve effort budget will typically
give rise to a reduction in overt performance under stress. We, therefore, posit the
following hypothesis:
H1: Exhaustion is positively associated with in-role performance problems.
Chapitre 5: Deuxième article
75
Contrary to employees experiencing gradual exhaustion whose capacity to
perform is reduced, employees experiencing gradual cynicism do not sustain an
impoverishment of their energy resources. Nevertheless, they are confronted with
insufficient resources to achieve their work goals. Because in-role performance
behaviors are monitored and reinforced by organizational rewards and sanctions
systems, allowing in-role performance to decline would result in direct negative
consequences for employees experiencing gradual cynicism (Bakker et al., 2004).
Hence, we argue that because they are confronted to a lack of resources, they will
choose to invest all of their energy resources in the challenging goal of meeting inrole performance standards - aiming to protect their self-interests - and not manifest
extra-role behaviors that directly benefit the organization. Under this strategy, which
corresponds with Hockey’s (1997) description of the active coping mode, effort will
remain well under the upper limit of the reserve effort budget although the level of
energy invested in primary task performance may be increased. Employees will
deliver minimal but sufficient job performance, i.e. they will meet in-role
performance standards. Consequently, the development of a cynical attitude should
not have a dramatic effect on in-role performance (which the cynical employee will
try to maintain), contrary to feelings of exhaustion, but should strongly inhibit OCBO.
Employees experiencing gradual cynicism under the influence of chronic
insufficient job resources, who have decided to participate in organizational activities
despite their health condition, will most likely be unmotivated to contribute to
organizational success and refrain from engaging in extra-role behaviors that directly
Chapitre 5: Deuxième article
76
benefit the organization. OCB-O is more likely to be associated with the motivational
process (cynicism) than the energetic process (exhaustion). As Bakker et al. (2004)
argue, when employees do possess job resources, they tend to go beyond actual work
goals accomplishment. Indeed, research from both the employer’s and employee’s
perspective has shown that an employee-organization relationship that reflects a
strong social exchange, which entails unspecified, broad, and open-ended obligations
by both parties, elicits a higher level of affective commitment to an employer and
more extra-role behaviors than a relationship characterized by mutual low obligations
or underinvestment by the employer (Shore & Barksdale, 1998; Tsui, Pearce, Porter,
& Tripoli, 1997). In a social exchange relationship, the employer offers inducements
that go beyond short-term monetary rewards, which denotes an extended
consideration of an employee’s well-being (Tsui et al., 1997). In return, the employee
is more supportive of organizational goals and is willing to make a contribution that
goes beyond manifesting in-role behaviors. However, when organizations do not
provide employees with sufficient job resources to meet high demands
(underinvestment by the employer), the likely consequences are low affective
organizational commitment and motivation to contribute to organizational success
that will be manifest in low extra-role performance.
Our reasoning is supported by studies which show that extra-role performance
in general and OCB-O are negatively and more strongly associated with cynicism
than with exhaustion, while exhaustion is a more robust predictor of in-role
performance in comparison to cynicism (e.g., Bakker et al., 2004; Halbesleben &
Chapitre 5: Deuxième article
77
Bowler, 2005). The influence of cynicism on in-role performance and OCB-O is
stated in the following hypotheses:
H2: Cynicism is positively associated with low extra-role performance/OCBO.
H3:
Exhaustion
is
less
strongly
associated
with
low
extra-role
performance/OCB-O than cynicism.
H4: Cynicism is less strongly associated with in-role performance problems
than exhaustion.
5.2.2.3 Burnout and absenteeism
March and Simon (1958) theorized that the decision to participate in
organizational activities is based on the notion of exchange. The individual balances
the inducements the organization provides and the contributions required to join,
attend and remain a member of the organization. Burning-out employees are likely to
perceive a lack of balance between the inducements the organization provides and the
contributions required from them when attending work. Because the job resources
provided are insufficient to meet the level of job demands, work attendance is likely
to result in health costs and losses of other resources that employees may consider
important (e.g., valued extrinsic rewards, work-related self-efficacy). Indeed, reduced
capacity and reduced motivation to work is indicative of resource losses. According
to March and Simon’s theory, employees who gradually enter a state of burnout will
have to decide at some point whether or not to participate in organizational activities.
Because of the perceived imbalance between inducements and contributions, burning-
Chapitre 5: Deuxième article
78
out employees are likely to fail to report for scheduled work, a phenomenon known
as absenteeism (Johns, 1995). Reasonable attendance can be viewed as a term in the
implicit contract between employees and employers and as a contribution in
employees’ self-constructed psychological contract (Johns, 2009). According to
Johns (2008), research concerning justice and contract breach constitutes some of the
best evidence for the strategic and proactive use of attendance as a means of
managing one’s relationship with the employer.
Because our research model emphasizes the choice burning-out employees
make between two alternative actions, i.e. taking a leave or continuing to go to work
even though they feel some degree of exhaustion or cynicism, we choose to focus on
short-term rather than long-term leaves of absence because the use of sick days once
in a while or for short periods of time is largely under the control of the individual.
Furthermore, the opportunity to exercise a choice in regard to participation in
organizational activities would be possible under conditions of moderate levels of
exhaustion and cynicism because a high degree of burnout would lead to involuntary
sickness absenteeism. Employees who reach this stage will most likely have sought
professional medical advice and be told to take a long-term leave of absence.
Although we argue that there exists a curvilinear relationship between burnout and
short-term leaves of absence in the form of an inverted-U shape curve, we do not
expect to observe a nonlinear effect in the present study because of range restriction.
Our sample only comprises employees still attending work and performing, therefore
experiencing a moderate degree of burnout at most.
Chapitre 5: Deuxième article
79
If burning-out employees can choose to use sickness absenteeism as a means
of restoring equity (Johns, 2009), by lowering their input in the exchange
relationship, short-term leaves of absence can help simultaneously restore their
physical and psychological equilibrium. Hence, short-term but frequent leaves of
absence as a response to feelings of exhaustion or cynicism could very well be
explained through a social exchange perspective and the restorative model of
absenteeism (cf. Johns, 2009). Sickness absenteeism as an equity restoring and
coping mechanism is well illustrated in the liberal use of the term “mental health
days” by nurses as a stated reason for absence in Hackett, Bycio, and Guion’s (1989)
idiographic study.
The empirical evidence demonstrates that absenteeism is one of the most
important organizational level consequences of burnout, although the effect appears
to be rather small and mostly related to exhaustion, as a recent meta-analysis shows
(cf. Darr & Johns, 2008). Furthermore, studies do not usually distinguish between
short-term and long-term leaves of absence. One notable exception is Firth and
Britton (1989), who report non-significant but positive relationships between the two
core dimensions of burnout and the number of short-term leaves of absence in the 12month period following the completion of the questionnaire. The non-significant
associations found may be due to the small size of their sample and to the fact that
these researchers have estimated bivariate relationships using Pearson's correlation
coefficients. Although no information was provided regarding the distribution of the
absence frequency measure used, distributional properties of absenteeism data
generally prohibit the use of such a conventional analysis (Martocchio & Harrison,
Chapitre 5: Deuxième article
80
1993). Violations of distributional assumptions attenuate the Pearson's correlations
(Carroll,1961, in Martocchio & Harrison, 1993). On the basis of the previous
arguments and empirical evidence, we therefore formulate the following hypotheses:
H5: Exhaustion increases the probability of future short-term leaves of absence.
H6: Cynicism increases the probability of future short-term leaves of absence.
5.2.2.4 Job performance and absenteeism
As suggested by Aronsson et al. (2000) and Johns (2008, 2009), it is possible
that low job performance, said to signal presenteeism, is a precursor of sickness
absenteeism. Individuals are thought to implement various coping strategies intended
to reduce stress and improve well-being (Edwards, 1992; Lazarus & Folkman, 1984),
with ineffective and less salient adjustment measures being replaced by more extreme
measures (Johns, 2009). Thus, when exposed to hindrance stressors, individuals may
first refrain from engaging in discretionary extra-role performance behaviours and try
to maintain their level of performance on the primary task by mobilizing additional
resources at the expense of an increase in energy costs, psychophysiological
activation and fatigue (cf. Hockey, 1997). When individuals experience chronic
stress, a process of depletion and wearing down of their energy resources develops
over time (Hobfoll & Shirom, 2000). This process fosters gradual psychological
detachment from work (i.e. cynicism), seen as a defensive adaptation strategy
centered on emotions intended to end a persistent state of exhaustion (Cherniss, 1980;
Leiter & Maslach, 1988; Shirom, 2003). As physical and psychological symptoms
escalate, individuals might expend greater efforts to cope such as problem solving or,
Chapitre 5: Deuxième article
81
because of existing constraints, exhibiting successively more salient withdrawal
behaviors ranging from reduced primary task performance to lateness, absence and
turnover (Darr & Johns, 2008; Johns, 2001). Research has provided strong evidence
for a continuum of withdrawal for some behaviors (Johns, 2001). However, because
studies conducted to date have mainly considered presenteeism and sickness
absenteeism as alternate or spillover forms of withdrawal behaviors (e.g., Aronsson et
al., 2000; Caverley, Cunningham, & Mac Gregor, 2007; Elstad & Vabo, 2008;
Johansson & Lundberg, 2004), evidence for a progression from presenteeism due to
burnout symptoms to sickness absenteeism is still lacking. However, a recent study
conducted by Bergström, Bodin, Hagberg, Aronsson and Josephson (2009) shows
that frequent episodes of presenteeism associated with various health problems, shortterm or chronic, may lead to future sickness absenteeism. Still, these researchers raise
the question whether presenteeism is a risk factor for sickness absenteeism regardless
of the specific health problem concerned. Furthermore, their study does not examine
the intermediary role that low job performance plays in the progression model of
withdrawal. Nevertheless, cross-sectional studies reveal that absenteeism is
negatively related to in-role and extra-role performance (cf. Johns, 2008). The
influence of low job performance ascribed to illness on future sickness absenteeism is
stated in the following hypotheses:
H7: In-role performance problems increase the probability of future short-term
leaves of absence.
H8: Low extra-role performance/OCB-O increases the probability of future
short-term leaves of absence.
Chapitre 5: Deuxième article
82
5.2.3 The boost effect of organizational vision and excellence values
Several plausible models, as discussed by Johns (2001), might describe the
connections between various withdrawal behaviors. Besides the progression model of
withdrawal that addresses the temporal nature of the relationship between
presenteeism and sickness absenteeism, the alternate forms model of withdrawal
considers the relationship between these variables at a particular point in time.
Johansson and Lundberg (2004) contend that presenteeism (termed sickness
attendance) can theoretically be considered the alternative action to being absent
because of ill health (termed sickness absence). Caverley et al. (2007) provide
evidence consistent with the substitution hypothesis. As Johns (2008) argues, the act
of presenteeism may occur when the option of absenteeism is not available or is
perceived as being too costly. Hence, as shown in the bottom part of Figure 1, our
research model emphasizes the choice burning-out employees must make between
two alternative actions, i.e. taking a leave or continuing to go to work even though
they feel some degree of exhaustion or cynicism. We propose that the articulation of
an organizational vision and the promotion of excellence values may motivate
burning-out employees to choose presenteeism over sickness absenteeism. Because
we had access to register data on sickness absences, we tested our hypotheses related
to the influence of organizational vision and excellence values on the decision to
maintain participation in organizational activities by using future sickness
absenteeism as a dependent variable. Thus, following Kivimäki et al. (2005),
burning-out employees who have not engaged in sickness absenteeism over the 12-
Chapitre 5: Deuxième article
83
month period following the measure of burnout symptoms are considered as having
chosen presenteeism.
There are at least two reasons why a well articulated organizational vision
could encourage presenteeism and, consequently, be associated with a lower
likelihood of sickness absenteeism. First, while it is true that burning-out employees
may perceive a lack of reciprocity from the organization, communicating an
organizational vision could sustain their contribution-reward expectancies. Vision is a
general transcendent ideal that can refer to a variety of desired outcomes (Kirkpatrick
and Locke, 1996). As such, it represents terminal values, i.e. a belief that a specific
end-state of existence is organizationally or even socially preferable (cf. Rokeach,
1973). Usually, this preferable end-state of existence is nothing less than excellence
(Testa, 1999). Kirkpatrick and Locke (1996) argue that vision, as a general and
challenging goal, may be effective because it directs attention toward desired
outcomes, thus arousing followers’ needs and values, and is highly discrepant from
the status quo but still within the realm of acceptance. Vision is used by effective
leaders to galvanize followers to put forth their best efforts toward a compelling goal
(Collins & Porras, 1991, in Testa, 1999), while instilling confidence in their ability to
reach that goal (Kirkpatrick and Locke, 1996). By providing a visual image of a
desirable yet possible future state of the organization, leaders are somehow promising
organizational members an eventual return on investment. Furthermore, by
emphasizing the strengths on which the organization can effectively capitalize upon
to materialize its vision, i.e. its employees’ unique characteristics and resources of
other kinds, leaders may favor a more positive perception among employees of their
Chapitre 5: Deuxième article
84
work environment and of their own personal resources. Thus, a well articulated
organizational vision is likely to develop positive psychological resource capacities
such as self-efficacy, hope, optimism and resiliency (cf. Luthans & Youssef, 2007)
that could enhance burning-out employees’ motivation to attend work.
A second reason why a well articulated organizational vision could encourage
presenteeism is that vision as a superordinate goal could also be thought as a form of
social control guiding members’ attitudes and behaviors (cf. O’Reilly and Chatman,
1996). When members agree and care about a common goal, violation of expectations
about appropriate attitudes and behaviors to achieve this goal may be sanctioned by
any member, the ultimate punishment being exclusion (O’Reilly and Chatman, 1996).
Vision is a challenging goal that commands dedication and high performance in all
organizational members. Vision broadens and elevates the interests of employees; it
stimulates them to look beyond their own self-interest for the good of the group
(Bass, 1990, in Testa, 1999). Hence, a well articulated organizational vision could be
associated with a lower likelihood of sickness absenteeism.
Emphasizing quality and innovation as organizational values may be another
way to incite burning-out employees to choose presenteeism over sickness
absenteeism. Quality and innovation represent instrumental values, i.e. a belief that
these two specific modes of conduct are organizationally or even socially preferable
(cf. Rokeach, 1973). These values are among the seven organizational values
identified by Peters and Waterman (1982) as representing excellence and also relate
to 2 of the 14 characteristics of magnet hospitals, which are known to be “institutions
of excellence” in the health care sector, namely quality of care and quality assurance
Chapitre 5: Deuxième article
85
(cf. Cook, Hiroz, & Mildon, 2006). Two reasons might explain why emphasizing
quality and innovation can promote presenteeism. First, O’Reilly and Chatman
(1996) contend that the power of organizational culture - to increase commitment
among members - may lie is the power of social control. As a system of shared values
defining what is important and norms defining appropriate attitudes and behaviors,
organizational culture is a vital component in the achievement of organizational goals
(Testa, 1999). Health care workers, especially those experiencing burnout symptoms,
are likely incline to believe that quality and innovation are desirable modes of
conduct. Struggling to meet performance standards in an unfavourable work
environment, they may conceive quality and innovation as preferable means to
improve patient’s well-being through better working conditions and practices for
them. When organizational members strongly believe that quality and innovation are
desirable modes of conduct, they are prone to put pressure on each other to attend,
especially under highly interdependent team structures (Barker, 1993, in Johns,
2008). Johns (2008) observes that presenteeism cultures have been described that
resemble absenteeism cultures, which are assumed to be underpinned by norms
dictating acceptable levels and patterns attendance and acceptable reasons for
absence. For instance, Simpson (1998) described a culture of “competitive
presenteeism” in a firm that had suffered downsizing. Deserved absence was avoided
to showcase one’s organizational commitment to senior managers. Thus, social
control can foster absence or attendance.
A second reason why emphasizing quality and innovation could encourage
presenteeism is that an organizational culture can play a significant role in the moral
Chapitre 5: Deuxième article
86
development of organizational members according to Kohlberg’s model of cognitive
moral development (1969). Even though burning-out employees may feel that their
contribution far exceeds the inducements provided by the organization, emphasizing
quality and innovation could stimulate them to look past their own personal
immediate interests and commit to the purpose of the group. A culture of excellence
based on values such as quality and innovation tends to heighten one’s sensitivity to
the concern of others, in this case those of the patient. It may make burning-out
employees’ own need for self-care appear less urgent and prompt them to adopt a
self-sacrificing behavior such as coming to work despite their health problems.
Organizational values are employed to engage people emotionally and provide them
with a sense of purpose; they fulfill people’s desire to be a part of valuable causes or
efforts (O’Reilly and Chatman, 1996). O’Reilly and Chatman (1996) assert that the
internalization of some organizational values such as helping others and contributing
to society can result in a perception of intrinsic value (that is something that the
person believes in rather than something imposed externally). Consequently, it may
be accompanied by more positive attitudes and chosen behaviors.
Following these arguments, we expect that the promotion of a culture of
excellence, by communicating an organizational vision and emphasizing quality and
innovation as organizational values, will enhance burning-out employees’ motivation
to attend work and encourage presenteeism rather than sickness absenteeism. Of
course, this boost effect of organizational vision and excellence values would not be
expected in the case of employees who reach a high degree of burnout. At this point,
the physical, mental and social aspects of well-being are severely affected and
Chapitre 5: Deuxième article
87
professional medical advice will most likely have been sought and a “doctor’s order”
given.
While we could not find any empirical research on the impact of
organizational vision and values on presenteeism/absenteeism, Boxx, Odom and
Dunn (1991) show that employees who work in a not-for-profit environment adhering
to the seven organizational values identified by Peters and Waterman (1982) as
representing excellence are more affectively committed to their organization.
Furthermore, Finegan (2000) concludes that the more an organization is perceived as
operating by values tapping into issues of “vision” (e.g., initiative, creativity), the
greater the employee’s affective organizational commitment. Hence, even though the
experience of burnout would tend to weaken the affective bond with the organization
(cf. Bakker, Demerouti, De Boer, & Schaufeli, 2003), promoting a culture of
excellence could help preserve this bond. According to Meyer and Herscovitch
(2001), a strong affective bond with the organization motivates employees to exert
effort to achieve organizational goals. One indication of this willingness to endeavor
to improve organizational effectiveness would be coming to work despite health
problems. Indeed, two meta-analyses have shown that affective organizational
commitment correlates positively with attendance (cf. Mathieu & Zajac, 1990;
Meyer, Stanley, Herscovitch, & Topolnytsky, 2002). On the basis of these arguments
and empirical evidence, we posit the following hypotheses:
H9: Perceived organizational vision and commitment to excellence values in
combination with exhaustion decrease the probability of future sickness
absenteeism (i.e. higher probability of future presenteeism).
Chapitre 5: Deuxième article
88
H10: Perceived organizational vision and commitment to excellence values in
combination with cynicism decrease the probability of future sickness
absenteeism (i.e. higher probability of future presenteeism).
5.3
Methodology
5.3.1 Participants
The study was carried out among workers of a large health and social services
center in the Canadian public health and social services sector. The organization
comprises a local community health center, a residential and long-term care center
and a general and specialized hospital. Questionnaires were distributed to all
employees of the establishment by a human resources professional. If the employees
were absent, the questionnaire was sent directly to their homes. In total, 545 (29%)
employees completed the questionnaire and 428 (79%) consented to their personal
absences file’s being matched to the questionnaire responses. The present study is
based on these 428 respondents, characterized as follows5: 88% are female, 41.5%
are between ages 30 and 45 (M=45), 65% are full-time employees, 82% work on the
day shift, 42% have been employed by the hospital for over 10 years (M=11) and
9.5% hold a management position.
5
Percentages reported are calculated based on valid cases.
Chapitre 5: Deuxième article
89
5.3.2 Measures
Unless otherwise specified, employees were asked to express their level of
agreement with each statement on a continuum ranging from Strongly disagree (1) to
Strongly agree (7).
Exhaustion (e.g., “I feel emotionally drained from my work”) and cynicism
(e.g., “I have become more cynical about whether my work contributes anything”)
were each measured using the five-item scale taken from the Maslach Burnout
Inventory – General Survey (MBI-GS) (Schaufeli et al., 1996).
In-role performance problems were assessed with a six-item scale taken from
the Health and Labour Questionnaire (Hakkaart & Essink, 2000). Employees who
did go to work over the past two weeks were asked to evaluate experienced
impediments at work due to health problems (i.e. concentration, working pace, need
to be alone, decision making, postponement of work and taking over work by other
workers) during this period. A sample item is “I had a problem concentrating.”
Responses range from (Almost) Never (1) to (Almost) Always (4).
The OCB-O scale comprises six items related to two dimensions: involvement
and initiative. One of the three items measuring involvement was taken from
Podsakoff, MacKenzie, Morman and Fetter’s (1990) civic virtue scale (e.g., “I attend
meetings that are not mandatory, but are considered important by the organization”);
the two other items were designed specifically for this study to tap active
participation in organizational activities (e.g., “I am active in department affairs”).
Chapitre 5: Deuxième article
90
The three items measuring initiative were taken from Morrison and Phelps’s (1999)
taking charge scale (e.g., “I try to implement solutions to pressing organizational
problems”). To maintain consistency in the formulation of our hypotheses pertaining
to the consequences of presenteeism, the scores for the six items were reversed and
combined to form a single measure of “low extra-role performance/OCB-O”,
following the results of the factor analysis to be discussed shortly.
Sickness
absenteeism
data
were
retrieved
from
the
organization’s
computerized registration system. We gathered absence data for the 12-month period
before and following the administration of the questionnaire. A period of one year
was chosen to ensure stability in the absence measures (Hammer & Landau, 1981).
Previous sickness absenteeism was used as a control variable in the present study
because it is considered to be the best predictor of future sickness absenteeism (cf.
Martocchio & Harrison, 1993). In this study, we were interested in one category of
absences specified by the organization: short-term leaves of absence that are illnessrelated. Absences that are less than five days in duration fall into the short-term
category and are not medically certified. Respondents were grouped into two
categories: 1) those who took short-term leaves of absence over the 12-month
following the administration of the questionnaire and 2) those who did not take shortterm leaves of absence over the same period.
Organizational vision was measured with three items derived from Oswald,
Mossholder and Harris’s (1994) vision salience scale and designed to gauge the
employee's degree of experienced vision salience. A sample item is “General
managers have a clear vision regarding the goals that we need to reach collectively.”
Chapitre 5: Deuxième article
91
Organizational values items were designed to capture the extent to which employees
believe their organization places importance on quality and innovation. Each value
was measured with three items derived from Van Dyne, Graham and Dienesch’s
(1994) instrument. Sample items are “High-quality services are a priority for my
organization” and “My organization strongly encourages its employees to be
innovative.” Following a factor analysis, discussed below, the scores for the nine
items were combined to form a single measure of perceived organizational vision and
commitment to excellence values, hereafter referred to as perceived “organizational
idealism.”
Control variables. In addition to previous sickness absenteeism (short-term
leaves of absence), information related to gender, age, marital status, number of
dependents and children under age 12, level of education, job status, work schedule,
organizational tenure, occupation and type of services offered by the health and social
services center (e.g., residential and long-term care) was collected in order to measure
the influence of these variables on the endogenous variables included in our model.
5.3.3 Analyses
We used two types of regression analyses to test our hypotheses, depending
on the nature of the dependent variable. Multiple regression analyses were performed
to test the hypotheses pertaining to in-role performance problems and low extra-role
performance/OCB-O (H1 to H4), while logistic regression analyses were conducted
to test the hypotheses pertaining to sickness absenteeism (H5 to H10). To examine
Chapitre 5: Deuxième article
92
interaction effects, we used Cohen, Cohen, West, and Aiken’s (2003) strategy of
forming a multiplicative term with the independent and moderator variables.
Following Jaccard, Turrisi and Wan’s (1990) recommendation, we first mean
centered the two predictors and then formed a product term from these transformed
scores. Multiple and logistic regression analyses were performed using hierarchically
formulated models in which the predictor variables were included in the following
order: step 1 - control variables and step 2 - focal independent variable. For
interaction effects, the moderator variable was also included at step 2 and the product
term at step 3.
5.4
Results
5.4.1 Preliminary analysis
Before testing our hypotheses, an exploratory factor analysis was first
conducted on the 25 scale items used to assess the 6 theoretical variables measured
with reflective indicators6 (i.e. the measures reflect the construct) considered in the
present study. Using an eigenvalue greater than the 1.00 cut-off criterion, three
factors emerged from the factor analysis. Factor loadings varied from .46 to .88 and
thus can be considered practically significant considering our sample size (cf. Hair,
Anderson, Tatham, & Black, 1998). The items were grouped as follows: 1)
exhaustion and cynicism, 2) organizational vision, quality and innovation, 3)
6
Given that the measures used to assess in-role performance problems are formative indicators (i.e. the
measures produce the construct), reliability in the internal consistency sense and construct validity in
terms of convergent and discriminant validity are not meaningful (Bagozzi, 1994).
Chapitre 5: Deuxième article
93
involvement and initiative. The factor solution provided a basis for the creation of
three summated scales. Two scales were labeled “low extra-role performance/OCBO”
(involvement
and
initiative)
and
“perceived
organizational
idealism”
(organizational vision and excellence values). These new constructs were used in
subsequent regression analyses for reason of parsimony. Whereas for the two core
dimensions of burnout, each subscale was retained because the JD-R model assumes
that two distinct processes contribute to the development of burnout, hence implying
that exhaustion and cynicism might have different consequences.
5.4.2 Hypothesis testing
Table I (p.94) presents the descriptive statistics, reliability coefficients and
correlations between the main variables of this study. As it can be seen from this
table, respondents in general are experiencing a moderate degree of burnout, as
reflected in average scores on the two subscales (exhaustion: M=3.72; cynicism:
M=3.12), which could be expected because our sample only comprises employees
still attending work and performing. Furthermore, Table I shows significant
correlations between the variables for which we have postulated a relationship.
Lastly, the level of reliability of internal consistency for the four variables measured
with multiple reflective indicators varies between .76 and .92, which satisfies the
minimum threshold of .70 generally recommended (cf. Hair et al., 1998).
Chapitre 5: Deuxième article
94
Table I
Descriptive statistics and correlations
Main variables
1. Exhaustion
2. Cynicism
3. In-role performance problems
4. Low extra-role performance/OCB-O
5. Future sickness absenteeism
6. Organizational idealism
M
SD
1
3.72
1.56
.91
3.12
1.46
.80**
.88
1.22
.45
.39**
.29**
--
2.77
.97
.19**
.28**
.16**
.76
.65
.48
.25**
.25**
.17**
.21**
--
4.50
1.15
-.40**
-.47**
-.25**
-.33**
-.26**
2
3
4
5
6
.92
Note. N=428. Pearson's correlation coefficients (r) are reported below the diagonal; coefficients alpha
( ) are reported on the diagonal. The reference category for the dichotomous variable sickness
absenteeism is “no short-term leave of absence”.
* p .05. ** p .01.
Table II (p.95) presents the results of the multiple regression analyses
performed to test the hypotheses pertaining to job performance (H1 to H4). In these
analyses, control variables that had a significant influence on the dependent variable
were included in the model. Hence, we controlled for previous sickness absenteeism
and job status when the dependent variable was in-role performance problems and for
previous sickness absenteeism, job status, work schedule and occupation when the
dependent variable was low extra-role performance/OCB-O. Job status (1=regular
full-time; 2=regular part-time, 3=on call), work schedule (1=day; 2=evening;
3=night; 4=rotation) and occupation (1=nursing and cardiorespiratory personnel;
2=paratechnical, auxiliary service and trade personnel; 3=office personnel; 4=health
and social services technicians; 5=health and social services professionals;
6=managers and professionals) are category variables for which dummy variables
have been created. The first category has been chosen as the reference category for all
the analyses.
Chapitre 5: Deuxième article
95
Table II
Results of the hierarchical multiple regression analyses
for in-role performance problems and low extra-role performance/OCB-O
Dependent
variable
Hypothesis 1
Hypothesis 2
Hypothesis 3
Hypothesis 4
In-role
performance
problems
Low
extra-role
performance/
OCB-O
Low
extra-role
performance/
OCB-O
In-role
performance
problems
Independent
variables
in the equation
Unstandardized
coefficients
B
S.E.
Standardized
coefficients
Beta
t
Sig.
95% C.I.
for B
Lower
Upper
Previous sickness
absenteeism
Job status 2
.01
.01
.05
1.00
.319
-.01
.02
-.09
.05
-.10
-2.05
.041
-.19
-.00
Job status 3
-.06
.10
-.03
-.56
.573
-.25
.14
Exhaustion
.12
.01
.39
8.57
.000
.09
.14
Previous sickness
absenteeism
Job status 2
.03
.02
.09
1.70
.091
-.01
.07
.14
.12
.06
1.19
.235
-.09
.37
Job status 3
.53
.26
.11
2.08
.039
.03
1.03
Work schedule 2
.23
.16
.07
1.41
.160
-.09
.55
Work schedule 3
.97
.30
.16
3.26
.001
.39
1.56
Work schedule 4
.24
.24
.05
1.04
.300
-.22
.71
Occupation 2
.27
.18
.08
1.52
.129
-.08
.61
Occupation 3
.05
.15
.02
.33
.745
-.24
.33
Occupation 4
.02
.19
.01
.11
.913
-.36
.40
Occupation 5
.08
.15
.03
.52
.607
-.22
.37
Occupation 6
-.56
.18
-.18
-3.05
.002
-.92
-.20
Cynicism
.14
.04
.21
4.04
.000
.07
.21
Previous sickness
absenteeism
Job status 2
.04
.02
.10
1.84
.066
-.00
.07
.15
.12
.07
1.25
.212
-.09
.39
Job status 3
.55
.26
.11
2.13
.034
.04
1.06
Work schedule 2
.23
.17
.07
1.37
.171
-.10
.56
Work schedule 3
.95
.30
.16
3.13
.002
.35
1.54
Work schedule 4
.28
.24
.06
1.17
.243
-.19
.75
Occupation 2
.20
.18
.06
1.13
.260
-.15
.55
Occupation 3
.01
.15
.00
.04
.966
-.28
.30
Occupation 4
.00
.20
.00
.01
.991
-.38
.39
Occupation 5
.05
.15
.02
.31
.756
-.25
.34
Occupation 6
-.64
.18
-.21
-3.52
.000
-1.01
-.28
Exhaustion
.08
.03
.12
2.29
.023
.01
.14
Previous sickness
absenteeism
Job status 2
.01
.01
.05
1.11
.269
-.01
.02
-.11
.05
-.11
-2.21
.027
-.20
-.01
Job status 3
-.10
.10
-.05
-.95
.341
-.30
.11
Cynicism
.09
.02
.28
6.00
.000
.06
.12
Note: Hypothesis 1: R2=.02 for Step 1; Δ R2=.15 for Step 2 (p<.001). Hypothesis 2: R2=.14 for Step 1;
Δ R2=.04 for Step 2 (p<.001). Hypothesis 3: R2=.14 for Step 1; Δ R2=.01 for Step 2 (p<.05).
Hypothesis 4: R2=.02 for Step 1; Δ R2=.08 for Step 2 (p<.001).
Chapitre 5: Deuxième article
96
Hypothesis 1 posited that exhaustion is positively associated with in-role
performance problems, while Hypothesis 2 stated that cynicism is positively
associated with low extra-role performance/OCB-O. Table II shows that both
hypotheses are confirmed, with exhaustion and cynicism being both positively
associated with their postulated dependent variable when the control variables are
included in the regression model. Also, Table II shows that exhaustion is more
strongly associated with in-role performance problems than cynicism, and that
cynicism is more strongly associated with low extra-role performance/OCB-O
than exhaustion, which confirms hypotheses 3 and 4. However, even though our
results reveal that cynicism is a better predictor of low extra-role performance/OCBO, we note that its relationship to in-role performance problems is even stronger (yet
remaining weaker than the relationship between exhaustion and in-role performance
problems).
Tables III (p.97), IV (p.98) and V (p.100) present the results of the logistic
regression analyses performed to test the hypotheses pertaining to sickness
absenteeism (H5 to H10). In these analyses, we controlled for previous sickness
absenteeism, age, job status and occupation because these variables had a significant
influence on the dependent variable. For logistic models with continuous variables
and no product term, the exponent of the logistic coefficient Exp(B) equals a
multiplicative factor (also called an odds ratio) by which the predicted odds change
given a 1-unit increase in the predictor variable, when all other predictor variables in
the equation are kept constant (Jaccard, 2001). For interactive logistic models with
two continuous predictors, X and Z, and a product term, XZ, the exponent of the
Chapitre 5: Deuxième article
97
Table III
Main effect of burnout core dimensions on future sickness absenteeism
Hypothesis 5
Dependent
variable
Independent
variables
in the equation
Future
sickness
absenteeism
Previous sickness
absenteeism
Age
Job status 2
Job status 3
Occupation 2
Occupation 3
Occupation 4
Occupation 5
Occupation 6
Exhaustion
Hypothesis 6
Future
sickness
absenteeism
Previous sickness
absenteeism
Age
Job status 2
Job status 3
Occupation 2
Occupation 3
Occupation 4
Occupation 5
Occupation 6
Cynicism
B
S.E.
Wald
df
Sig.
95% C.I.
for EXP(B)
Exp(B)
or
odds ratio
Lower
Upper
.52
.09
36.17
1
.000
1.68
1.42
1.99
-.03
.02
2.92
1
.087
.98
.95
1.00
-.32
.28
1.30
1
.255
.73
.42
1.26
.41
.59
.48
1
.488
1.51
.47
4.80
-.30
.45
.45
1
.502
.74
.31
1.79
-.97
.37
7.00
1
.008
.38
.19
.78
-.58
.52
1.22
1
.269
.56
.20
1.56
-.32
.35
.84
1
.361
.73
.37
1.44
-3.12
.65
22.75
1
.000
.04
.01
.16
.33
.09
13.66
1
.000
1.39
1.17
1.66
.51
.09
35.48
1
.000
1.66
1.41
1.97
-.02
.02
2.57
1
.109
.98
.95
1.01
-.34
.28
1.54
1
.214
.71
.41
1.22
.35
.60
.34
1
.561
1.41
.44
4.54
-.23
.45
.26
1
.613
.80
.33
1.92
-.92
.36
6.37
1
.012
.40
.20
.82
-.41
.52
.61
1
.433
.67
.24
1.84
-.29
.35
.72
1
.397
.75
.38
1.47
-3.02
.67
20.39
1
.000
.05
.01
.18
.30
.10
9.46
1
.002
1.35
1.11
1.63
logistic coefficient Exp(B) of the product term is the multiplicative factor by which
the multiplicative factor Exp(B) of X changes given a 1-unit increase in Z (Jaccard,
2001). The confidence interval for this parameter provides a sense of the sampling
error.
Chapitre 5: Deuxième article
98
Table IV
Main effect of types of job performance on future sickness absenteeism
Hypothesis 7
Hypothesis 8
Dependent
variable
Independent
variables
in the equation
Future
sickness
absenteeism
Previous sickness
absenteeism
Age
Job status 2
Future
sickness
absenteeism
B
S.E.
Wald
df
Sig.
.51
.08
36.94
1
.000
Exp(B)
or
odds ratio
1.66
-.03
.02
3.23
1
.073
-.27
.28
.95
1
.331
95% C.I.
for EXP(B)
Lower
Upper
1.41
1.96
.97
.95
1.00
.76
.44
1.31
Job status 3
.33
.59
.31
1
.576
1.39
.44
4.45
Occupation 2
-.41
.45
.83
1
.361
.67
.28
1.60
Occupation 3
-.92
.36
6.52
1
.011
.40
.20
.81
Occupation 4
-.36
.52
.48
1
.488
.70
.25
1.93
Occupation 5
-.43
.35
1.53
1
.217
.65
.33
1.29
Occupation 6
-3.20
.65
24.24
1
.000
.04
.01
.15
.93
.37
6.25
1
.012
2.53
1.22
5.22
In-role
performance
problems
Previous sickness
absenteeism
Age
.51
.08
36.83
1
.000
1.67
1.41
1.96
-.03
.02
3.53
1
.060
.97
.95
1.00
Job status 2
-.40
.27
2.14
1
.144
.67
.39
1.15
Job status 3
.15
.59
.06
1
.803
1.16
.36
3.70
Occupation 2
-.33
.45
.55
1
.459
.72
.30
1.73
Occupation 3
-.98
.36
7.38
1
.007
.38
.19
.76
Occupation 4
-.31
.52
.35
1
.554
.73
.26
2.05
Occupation 5
-.38
.34
1.21
1
.271
.69
.35
1.34
Occupation 6
-3.01
.67
20.48
1
.000
.05
.01
.18
Low extra-role
performance/
OCB-O
.30
.14
4.65
1
.031
1.35
1.03
1.77
Hypotheses 5 and 6 posited that each burnout dimension increases the
probability of future short-term leaves of absence. Table III (p.97) shows that a 1-unit
change in exhaustion increases the likelihood of future short-term leaves of absence
by 39% when the control variables are included in the regression model, compared
with an increase of 35% for cynicism, which supports Hypotheses 5 and 6.
Hypotheses 7 and 8 stated that in-role performance problems and low extrarole performance/OCB-O increase the probability of future short-term leaves of
Chapitre 5: Deuxième article
99
absence. Again, Table IV (p.98) shows that a 1-unit change in role-performance
problems increases the likelihood of future short-term leaves of absence by 153%
when the control variables are included in the regression model. However, we note
that the confidence interval for the exponent of the logistic coefficient of in-role
performance problems is remarkably large: the lower and upper limits are 1.22 and
5.22. In contrast, Table IV shows that a 1-unit change in low extra-role
performance/OCB-O increases the likelihood of future short-term leaves of absence
by only 35% when the influence of control variables is considered in the regression
model. Overall, these results confirm Hypotheses 7 and 8.
Hypothesis 9 posited that perceived organizational vision and commitment to
excellence values, in combination with exhaustion, decrease the probability of future
sickness absenteeism (i.e. higher probability of future presenteeism). Table V (p.100)
shows that the influence of the product term “exhaustion X organizational idealism”
is marginally significant at the standard 5% error rate when the control variables are
included in the regression model. Indeed, the confidence interval for the exponent of
the logistic coefficient contains the value of 1.0. However, given the difficulty of
detecting interaction effects in field studies, it is common practice to use a relaxed
Type 1 error rate (McClelland & Judd, 1993). Thus, if we raise the error rate to 10%,
the lower and upper limits of the confidence interval for the exponent of the logistic
coefficient of the product term become .73 and .98. The interpretation of the
significant interaction effect at the 10% error rate is as follows. The exponent of the
logistic coefficient is .85, meaning that a 1-unit increase in perceived organizational
idealism
reduces
the
influence
of
exhaustion
on
future
sickness
Chapitre 5: Deuxième article
100
Table V
Interaction effect between burnout core dimensions and organizational idealism
on future sickness absenteeism
Dependent
variable
Hypothesis 9
Future
sickness
absenteeism
Independent
variables
in the equation
Previous sickness
absenteeism
Age
Job status 2
Job status 3
Occupation 2
Occupation 3
Occupation 4
Occupation 5
Occupation 6
B
S.E.
Wald
df
Sig.
Exp(B)
or
odds
ratio
95% C.I.
for EXP(B)
Lower
Upper
.50
.09
34.17
1
.000
1.65
1.40
1.96
-.03
.02
3.36
1
.067
.97
.95
1.00
-.40
.28
1.95
1
.163
.67
.39
1.17
.41
.59
.49
1
.485
1.51
.48
4.80
-.31
.46
.46
1
.500
.74
.30
1.80
-.88
.37
5.67
1
.017
.42
.20
.86
-.57
.54
1.12
1
.290
.56
.20
1.63
-.28
.35
.63
1
.426
.76
.38
1.50
-3.00
.67
19.62
1
.000
.05
.01
.19
.27
.10
8.23
1
.004
1.32
1.09
1.59
-.28
.15
3.66
1
.056
.76
.57
1.01
-.17
.09
3.49
1
.062
.85
.71
1.01
.50
.09
33.09
1
.000
1.64
1.39
1.94
-.02
.02
2.61
1
.107
.98
.95
1.01
-.37
.28
1.77
1
.183
.69
.40
1.19
.36
.60
.36
1
.549
1.43
.44
4.61
-.25
.46
.31
1
.580
.78
.32
1.90
-.83
.37
5.08
1
.024
.44
.21
.90
-.40
.53
.56
1
.455
.67
.24
1.91
-.28
.35
.66
1
.417
.75
.38
1.50
-3.00
.69
18.36
1
.000
.05
.01
.20
.24
.11
5.24
1
.022
1.27
1.04
1.56
-.29
.15
3.97
1
.046
.75
.56
1.00
-.22
.11
4.20
1
.041
.81
.66
.99
Exhaustion
Organizational
idealism
Hypothesis 10
Future
sickness
absenteeism
Exhaustion X
Organizational
idealism
Previous sickness
absenteeism
Age
Job status 2
Job status 3
Occupation 2
Occupation 3
Occupation 4
Occupation 5
Occupation 6
Cynicism
Organizational
idealism
Cynicism X
Organizational
idealism
Chapitre 5: Deuxième article
101
absenteeism by a factor of .85. Given that the exponent of the logistic coefficient of
exhaustion equals 1.32, if we increase perceived organizational idealism by one unit,
then the odds ratio for future sickness absenteeism decreases by a factor of .85, from
1.32 to 1.12 (1.32 x .85 = 1.12). If we increase perceived organizational idealism by
another unit, the odds ratio for future sickness absenteeism is reduced to below the
critical value of 1.00 and reaches .95 (1.12 x .85 = .95).
Hypothesis 10 posited that perceived organizational vision and commitment to
excellence values, in combination with cynicism, decrease the probability of future
sickness absenteeism (i.e. higher probability of future presenteeism). Table V shows
that the influence of the product term “cynicism X organizational idealism” is
significant when the control variables are included in the regression model, which
confirms Hypothesis 10. The exponent of the logistic coefficient is .81, meaning that
if we increase perceived organizational idealism by one unit, then the odds ratio for
future sickness absenteeism decreases from 1.27 to 1.03. If we increase perceived
organizational idealism by another unit, the odds ratio for future sickness absenteeism
is reduced to below the critical value of 1.00 and reaches .83.
5.5
Discussion
The present study investigated whether, as suggested by Aronsson et al.
(2000) and Johns (2008, 2009), there is a risk that the sickness presenteeism of today
will become the sickness absenteeism of tomorrow. Evidence for a progression from
presenteeism due to burnout symptoms to sickness absenteeism was still lacking.
Chapitre 5: Deuxième article
102
Indeed, earlier studies mainly considered presenteeism and sickness absenteeism as
alternate or spillover forms of withdrawal behaviors (e.g., Aronsson et al., 2000;
Caverley et al., 2007; Elstad & Vabo, 2008; Johansson & Lundberg, 2004) or did not
examine the impact of presenteeism on future absence behavior by taking into
account the specific health problem concerned (cf. Bergström et al., 2009).
Furthermore, in contrast with presenteeism research conducted to date, we adopted a
broader perspective by examining how a low level of overall job performance including low extra-role performance - due to the gradual intensification of burnout
symptoms among employees who choose to bring themselves to work despite their
ill-health, an act known as presenteeism, increases the probability of future sickness
absenteeism. In addition, the present study contributes to the understanding of the
factors that influence a person’s choice between presenteeism and sickness
absenteeism, two alternative types of illness behaviors. We examined the influence of
macro-organizational factors characterizing a culture of excellence, namely vision
and values, on the decision to participate in organizational activities. Such a culture
has been criticized for making the achievement of firm goals the only outcome that
matters and for compromising employee’s health (e.g., Aubert & De Gaulejac, 1991;
Godard & Delaney, 2000).
5.5.1 Interpretation of results
The results of the present study suggest that burning-out employees who
decide to attend work, albeit at a reduced capacity or with reduced motivation
because of feelings of exhaustion or cynicism, encounter more in-role performance
Chapitre 5: Deuxième article
103
problems or are more likely to refrain from engaging in extra-role behaviors that
directly benefit the organization. These findings corroborate results of other studies
that have shown that exhaustion is negatively associated with in-role performance (cf.
Taris, 2006) and that extra-role performance in general and OCB-O are negatively
associated with cynicism (e.g., Bakker et al., 2004; Halbesleben & Bowler, 2005).
Also, as reported in other studies (e.g., Bakker et al., 2004; Halbesleben & Bowler,
2005), we found that exhaustion is more strongly associated with in-role performance
problems than cynicism and that cynicism is a more robust predictor of low extra-role
performance/OCB-O in comparison to exhaustion. However, we observed that
cynicism is more strongly associated with in-role performance problems than with
low extra-role performance/OCB-O. This unexpected finding contradicts our
reasoning. We asserted that the development of a cynical attitude should not have a
dramatic effect on in-role performance which the cynical employee will try to
maintain because in-role performance behaviors are monitored and reinforced by
organizational rewards and sanctions systems. However, we argued that the
development
of
a
cynical
attitude
should
strongly
inhibit
extra-role
performance/OCB-O. In their study, Bakker et al. (2004) have indeed found that
cynicism better predicts extra-role performance in general (a composite measure of
OCB-O and OCB-I) than in-role performance, both assessed by one colleague. A
possible explanation for these contradictory results is given by Halbesleben and
Bowler’s (2005) study in which different types of job performance were assessed by
the participant, his supervisor and his closest coworker. Cynicism was found to be
more strongly related to OCB-O than to in-role performance when job performance
was assessed by a peer. When job performance was assessed by the participant or the
Chapitre 5: Deuxième article
104
supervisor, the relationships between cynicism and the two types of job performance
were similar in strength. Interestingly, results for OCB-I - no matter who rated job
performance - followed the pattern of relationships observed by Bakker et al. (2004):
cynicism was always more strongly associated to OCB-I than to in-role performance.
Perhaps colleagues can assess more accurately OCB-O than supervisors because they
are more likely to confide in each other, and because they have more opportunities to
observe each other behaviors (but not necessarily OCB-I, as these behaviors can be
directed toward supervisors). Coworkers may also be in a better position to assess
accurately OCB-O than burning-out employees themselves because their perception
is not biased by negative affect toward an organization perceived as uncaring (but not
necessarily OCB-I, as these behaviors are intended to benefit directly specific
individuals and not the organization).
Furthermore, we note that the relationship between exhaustion and in-role
performance problems is relatively stronger than the relationship between cynicism
and low extra-role performance/OCB-O. These findings, together with the
examination of the strength of the bivariate relationships between burnout symptoms
and the two types of job performance, lead us to conclude that the most important
consequence for health care organizations when burning-out employees have decided
to “go present” is a decrement in in-role performance, mainly due to experienced
feelings of exhaustion. Also, the present study provides strong support for the
proposition that low job performance said to signal presenteeism is a precursor of
sickness absenteeism and adds to the existing knowledge on a continuum of
withdrawal for some behaviors (Johns, 2001). The results indicate that in-role
Chapitre 5: Deuxième article
105
performance problems and low extra-role performance/OCB-O increase the
probability that employees will choose sickness absenteeism over presenteeism in the
following year. However, in-role performance problems, mainly associated with
experienced feelings of exhaustion, were found to be a stronger predictor of the
likelihood of choosing to take short-term leaves of absence. Therefore, burning-out
employees manifesting low extra-role performance/OCB-O are more likely than
those with in-role performance problems to choose to “go present.”
Finally, the results reveal that the articulation of an organizational vision and
the promotion of excellence values as perceived by employees moderate the
relationship between burnout symptoms and future sickness absenteeism. This study
shows that perceived organizational idealism prompt burning-out employees to
choose presenteeism over sickness absenteeism in the following year. This finding is
consistent with prior evidence that employees working for organizations that they
perceived as adhering to values tapping into issues of “vision” and representing
excellence are more affectively committed to their organization (Boxx et al., 1991;
Finegan, 2000). The influence of the moderator variable was found to be comparable
in magnitude whether the employee is experiencing feelings of exhaustion or
cynicism. Nevertheless, stronger evidence was found for a boost effect among
employees who develop a cynical attitude. This is not surprising because feelings of
exhaustion are more likely to engender physical health problems that may restrain
employees’ capacity to exercise a choice in regard to participation in organizational
activities.
Chapitre 5: Deuxième article
106
5.5.2 Study limitations
This study has some limitations pertaining, for instance, to its design. First,
the sample does not allow strong inferences because it represents one single
organization , and we could not examine non-response bias. Also, except for sickness
absenteeism, data on the other variables of our research model were obtained through
self-reported measures administrated at a particular point in time, a form of data
collection that might have introduced a problem of common method variance.
Although Podsakoff and Organ (1986) argue that the results of Harman’s one-factor
test cannot be interpreted unequivocally, we provided with the factor analysis some
evidence that common method variance does not account for most of the relationship
between cynicism and low extra-role performance/OCB-O because the items used to
assess these variables did not tend to load on a single general factor. Whereas for inrole performance problems, the measures used were formative indicators, which are
less likely to induce a desire on the part of the respondents to be consistent across
items and sections of a survey. In future studies, low job performance ascribed to
illness should be assessed with more objective measures (i.e. others’ ratings,
quantitative data).
Some limitations concern some of the measures used. In-role performance
problems were assessed by asking employees to evaluate experienced impediments at
work due to health problems. The instrument that was used is well-known to
researchers who portray presenteeism as the individual productivity loss associated
with the behavior of attending work despite health problems (cf. Mattke et al., 2007).
Chapitre 5: Deuxième article
107
As we argued, the behavior of attending work with health problems, i.e.
presenteeism, should not be confounded with the potential consequence of such a
behavior in terms of decreased job performance. Yet, asking about in-role
performance problems encountered because of health problems might have inflated
the strength of the relationship between feelings of exhaustion among employees
attending work and in-role performance problems. Nevertheless, feelings of
exhaustion and health problems are not exactly the same construct, the latter
conveying a more inclusive sense. Moreover, we acknowledge that the measure used
to assess in-role performance problems might have introduced a problem in terms of
construct validity. Although the relationships with the other variables, based on a
priori set hypotheses, were confirmed, the scale consists of formative indicators that
reflect different potential experienced impediments at work that may not be
exhaustive nor relevant for all workers, depending on their specific occupation and
status level (e.g., concentration, decision making, postponement of work). In
addition, although a strength of this study is the use of register data on sickness
absences, seeing the word sickness in front of the word absences is no guarantee that
the portion of absenteeism under consideration is truly a function of a verifiable
medical condition (Johns, 2009). Short-term sickness absences, because they are
largely “self-certified”, can have other causes than ill health, both “legitimate” (e.g.,
life crises) and “illegitimate” (e.g., to be off sick to play golf) (Johansson &
Lundberg, 2004). Hence, our measure of short-term sickness absenteeism cannot be
considered the perfect alternative action of showing up at work with health problems
(i.e. presenteeism).
Chapitre 5: Deuxième article
108
5.5.3 Practical implications
Although the above mentioned limitations preclude strong conclusions, the
results of the present study suggest that if the most important consequence for health
care organizations when burning-out employees have decided to “go present” is a
decrement in in-role performance mainly ascribed to the exhaustion dimension of the
burnout syndrome, employees experiencing feelings of exhaustion are also more
likely to “go absent.” Therefore, health care organizations are not only susceptible to
insufficient staffing and a high work pace that could trigger symptoms of exhaustion
among workers still attending their job, but could also contend with cynical
employees who are less likely to “go absent”.
One might be inclined to think that the presence at work of exhausted
employees is far more detrimental to their own health and the well-being of patients,
who could be exposed to fatigued, inattentive or even contagious staff (Elstad &
Vabo, 2008), than the presence at work of cynical employees. However, the
motivation to work of employees experiencing feelings of cynicism is far more
affected, because of the motivational process that originates from resources, in
comparison with the motivation to work of employees experiencing feelings of
exhaustion, whose capacity to work is more of a concern. Harrison, Johns and
Martocchio (2000) submit that there are a number of trends in the external
environment of contemporary organizations that have highlighted the need for high
service quality and the role of employees as a critical strategic resource in the
attainment of this challenging goal. As they argue, isolated and narrow views of one’s
Chapitre 5: Deuxième article
109
job responsibilities are not conducive to the kind of teamwork and innovation that are
necessary to achieve such a goal. The quest for excellence in which organizations are
engaging calls for more flexible and inclusive role orientations by employees. Yet, as
the present study shows, cynical employees, who are less likely to “go absent,” will
also be unmotivated to contribute to organizational success and restrain themselves
from engaging in extra-role behaviors that directly improve organizational
effectiveness. However, the results also indicate that the articulation of an
organizational vision and the promotion of excellence values by these organizations,
if perceived as such by burning-out employees, will be a stronger force for improved
attendance among employees experiencing feelings of cynicism than among those
experiencing feelings of exhaustion. Hence, cynical employees will be more incline
to attend work at an organization perceived as pursuing excellence and, as a
speculative conclusion, somehow more willing to consider extra-role behaviors as a
role requirement. However, these more predisposed employees would not be able to
convert the theory of dictating quality into action, unless they have access to
additional resources. Insufficient job resources are the main reason they have
developed a cynical attitude. It is when employees possess job resources that they
tend to go beyond actual work goal accomplishment (Bakker et al., 2004).
Finally, perceived organization vision and values have been depicted as
macro-organizational factors characterizing a culture of excellence that can enhance
burning-out employees’ motivation to attend work. The postulated motivational role
played by vision and values would make these work environment characteristics job
resources. Furthermore, Demerouti, Le Blanc, Bakker, Schaufeli and Hox (2009)
Chapitre 5: Deuxième article
110
argue that presenteeism can be viewed more positively as a sign of high commitment
and a type of organizational citizenship behavior. However, managers and
occupational health professionals should be aware that presenteeism is a choice that
will ultimately be counterproductive as it is more likely to further damage well-being
than to stimulate it (cf. Demerouti et al., 2009; Johns, 2009). Hence, this means that
some job resources, in the long term, could also impact workers, organizations and
their clients in an unbeneficial way.
5.5.4 Research avenues
In the present study, we chose to examine the impact of burnout when
employees attend work on job performance, and how low job performance said to
signal presenteeism is a precursor of sickness absenteeism. This approach implied
that burning-out employees could have taken sick leaves over this period during
which job performance was assessed. We treated previous absenteeism as a control
variable because we were interested in examining the consequences of showing up at
work with health problems on job performance and future sickness absenteeism. One
worthy research avenue would be to examine the impact of burnout symptoms on job
performance and future sickness absenteeism when employees choose not to take any
sick leave in the preceding year. The investigation of this matter in the present study
would have necessitated a larger sub-sample of respondents who took no short-term
leaves of absence over the 12-month period before the administration of the
questionnaire. Nevertheless, we presume that burning-out employees who have never
missed a day of work are more prone to show a lower level of overall job
Chapitre 5: Deuxième article
111
performance and, therefore, to choose sickness absenteeism over presenteeism in the
following year. Yet, having not taken any sick leave for a year, these burning-out
employees may rather be more likely not to take any short-term leaves of absence in
the following year and be given instead a medically certified long-term sickness
absence, as results from a qualitative suggest (cf. Grinyer and Singleton, 2000).
More research is also needed on other work environment factors known to be
associated with a culture of excellence, such as transformational leadership and highinvolvement management practices (i.e. training, information sharing, non-monetary
recognition), but whose relations with presenteeism have not been investigated.
Furthermore, it would be worthwhile to examine the relative influence of perceived
organizational commitment to humanistic values such as consideration and
acceptance of one another’s limits on burning-out employees’ decision to participate
in organizational activities.
Finally, as Elstad and Vabo’s (2008) study findings suggest, job stress,
sickness absenteeism and presenteeism may follow a certain dynamic such that
sickness absenteeism might also precede presenteeism. As the authors explain, in
organizations with relatively scarce staff resources such as health care establishments
(e.g., because of restricted funding, inability to recruit or high rate of absenteeism),
increased job stress elevates the rate of sickness absence. When staff resources
become even scarcer and job stress increases further, however, the care burden may
be increasingly perceived as inflexible, and further reductions in service delivery may
be viewed as intolerable by care workers because of professional norms and moral
obligations. However, because Elstad and Vabo considered sickness absenteeism and
Chapitre 5: Deuxième article
112
presenteeism as alternative types of illness behavior in their study, evidence for a
progression from sickness absenteeism to presenteeism is still lacking. Is it indeed
possible that at some critical point, burning-out employees might think that going to
work at the risk of offering a sub-optimal performance, perhaps a counter-productive
performance in some respects, is better for the patients than being absent and offering
no performance at all? An interesting question that follows would then be if their
perception of a strong organizational vision and commitment toward excellence
values would incite them to use a particular strategy or resource, such as social
support from colleagues, in order to provide high quality care in spite of their reduced
capacity and motivation to work.
CHAPITRE 6: TROISIÈME ARTICLE
Burnout among nursing staff and intention to leave the profession:
Validation of the job demands-resources model
Geneviève Jourdain and Denis Chênevert7
Abstract
The aim of the paper is to look at the relationships between stress factors related to
nurses’ work and social environment, burnout and intention to leave the nursing
profession. The research model was tested on cross-sectional data from 1,636
registered nurses working in hospitals who responded to a self-administrated
questionnaire. Results from our study suggest that burnout plays an important role in
nurses' decision to leave their profession, through its impact on psychosomatic
complaints and professional commitment. In order to retain nurses within the
profession, results indicate that nurses’ tasks and role should be restructured to reduce
work overload and increase the meaning of their work.
7
GJ reviewed the literature, conceived the research model and the questionnaire, analyzed and
interpreted the data and wrote the paper. DC supervised the research project, helped with the mailing
of the questionnaire and commented on the questionnaire and the manuscript.
Chapitre 6: Troisième article
6.1
114
Introduction
Nurses’ level of burnout and how it may lead to withdrawal manifestations is
especially important in the current context of nurses shortage. To date, much of the
research on burnout has investigated its links with milder (e.g., within-organizational
moves), serious (e.g., absenteeism) and severe forms of withdrawal (e.g.,
organizational exit). In contrast, the process whereby burnout prompts a career
change is poorly understood. There is little evidence about what the level of
occupational turnover among the nursing workforce might be and how it might differ
across countries. In Canada, between 2003 and 2007, more than 6% of employed
registered nurses (RNs) did not renewed their licence the following year in the same
province or territory (Canadian Institute for Health Information, 2008). The RNs in
the age group 40 to 49 demonstrated the lowest “exit” rates between 2003 and 2007
(2.2% on average annually). Higher “exit” rates were seen in both the youngest (a
mean of 6.7% for RNs under the age of 30) and the oldest age groups (a mean of 14%
for RNs over the age of 60). The average annual “exit” rate for RNs in the age groups
30 to 39 and 50 to 59 was 4.5%. It should be noted that an “exit” might be an RN
who is still practising nursing in another province, territory or country, or an RN who
has retired or left the profession (either permanently or temporarily). However, the
report states that many RNs who take a leave of absence or pursue further education
maintain their registration and are thus not counted as “exits”.
Chapitre 6: Troisième article
115
In light of current concerns over nursing shortages, nurses intent or decision to
leave the profession is a topic of great importance. Furthermore, occupational
turnover among nurses may discourage prospective students to join the profession.
According to Simoens, Villeneuve and Hurst (2005), current nurse shortages seem to
be caused, among other factors, by fewer young people entering the workforce, the
low social value given to nursing, and negative perceptions of nurse working
conditions. Evidence gathered to date indicates that it is less the nursing work per se
that motivates the decision to leave the profession than the conditions under which
the work takes place (Cheung, 2004; Wandelt, Pierce, & Widdowson, 1981). While
we know that many of these conditions that prompt nurses to leave their profession
are also stress factors commonly associated with burnout, no study has properly
investigated the links between stress factors, burnout and the decision to leave the
nursing profession. The present study employs self-reported intention to leave
nursing as the dependent variable. On the basis of Fishbein and Ajzen’s (1975) theory
of reasoned action and empirical support for an association between turnover
intention and actual turnover behaviour (Griffeth, Hom, & Gaertner, 2000), it is
reasonable to assume that some action on the part of the employee is likely to follow
self-expressed intention. Consequently, the present study has two aims. The first is to
examine the role of burnout in the relationship between stress factors related to
nurses’ environment and intention to leave the profession. The second aim is to
investigate the nature of the relationship between burnout and intention to leave the
nursing profession. Our intent is to make a contribution to the literature on career
change by unravelling the role of health in the process associated with the intention to
change career. Models on intention to change careers (Rhodes & Doering, 1983),
Chapitre 6: Troisième article
116
career motivation (London, 1983) and career commitment (Aryee & Tan, 1992)
include a host of situational, attitudinal and individual predictors but do not consider
health problems.
In addition, this study intends to shed light on the way in which stress factors
affect the development of burnout by investigating the role of factors that have rarely
been empirically evaluated: empowerment, as a psychological resource linked to
work, along with recognition and hostility from physicians and patients, two stress
factors associated with nurses’ social environment. By targeting the distinctive effect
of the factors studied by source, we attempt to identify the causes of the irritating
situations, i.e. the supervisor, colleagues or physicians, and to better understand the
role of various players in the burnout process leading to the intention to leave the
profession.
6.2
Theoretical Framework and Hypotheses
6.2.1 Definition of Burnout
Maslach and Jackson (1996) define burnout afflicting workers in human
services institutions as “a syndrome of emotional exhaustion, depersonalization, and
reduced personal accomplishment that can occur among individuals who work with
people in some capacity” (p.4). This meta-concept comprises three dimensions: 1)
emotional exhaustion refers to feelings of being emotionally drained and exhausted
(either physically or cognitively) by one’s work; 2) depersonalization is characterized
Chapitre 6: Troisième article
117
by feelings of insensitivity and impersonal responses toward clients; and 3) lack of
personal accomplishment refers to feelings of incompetence and unsuccessful
achievement in one’s work with people. Although the multidimensional nature of this
construct has been vigorously debated, many authors advocate a two-dimensional
concept that includes the emotional exhaustion and depersonalization components
(e.g., Büssing & Glaser, 2000; Demerouti et al., 2001; Green, Walkey, & Taylor,
1991; Kalliath, O'Driscoll, Gillespie, & Bluedorn, 2000). These authors argue that the
specificity of the syndrome lies in the combination of general reactions linked to
stress - captured by the emotional exhaustion dimension - and specific attitudinal
manifestations that signal a crisis in the individual-work relationship (Maslach,
Schaufeli, & Leiter, 2001). On the basis of their arguments and because this study
intends to validate the job demands-resources model, which adopts this twodimensional definition of burnout, we have excluded the third component from our
definition.
6.2.2 The Job Demands-Resources Model
As Figure 1 (p.118) illustrates, this study uses the job demands-resources (JDR) model (Demerouti et al., 2001) to clarify the role of burnout among nursing staff
in the relationship between stress factors and intention to leave the profession. The
model embraces the dominant theoretical perspective, put forth by Lazarus and
Folkman (1984), whereby burnout is a persistent dysfunctional state that results from
prolonged exposure to chronic stress, that is a situation where a person feels
incessantly confronted with a high level of demands and insufficient resources linked
Chapitre 6: Troisième article
118
to work itself and to the context in which the work takes place (Buunk et al., 1998;
Cooper et al., 2001; Leiter & Maslach, 2004).
Figure 1
Research model
Ressources
Resources
(latent exogenous variables)
Competence
Decision-making
Decision making authority
Meaning
Meaning
Impact
Support from supervisor
Support
Support from
from colleagues
colleagues
Recognition by physicians
Recognition
by patients
médecins
Reconnaissance des patients
Motivational process
H5 (a-h)
H7
H9
H8
H2 (a-h)
H4
Demands
(latent exogenous variables)
Quantitative overload
Role stress
Work interference with family
Hostility from physicians
Hostility from patients
Professional
commitment
Depersonalization
Intention to leave
the profession
H6
H10
Emotional
exhaustion
H1(a-e)
Psychosomatic
complaints
H3
Energetic process
The expression “stress factors” therefore refers both to the presence of
negative conditions (demands) and to the absence of positive conditions (resources).
Demands refer to those psychological, social, organizational and physical aspects of
the job that require an effort by the employee (Bakker et al., 2004). Resources refer to
those psychological, social, organizational and physical aspects of the job that are
intended to facilitate the attainment of work goals, to stimulate personal growth or to
reduce job demands (Bakker et al., 2004). Various theoretical approaches to burnout
have been proposed, ranging from individual to societal (cf. Schaufeli and Enzmann,
1998). These approaches, not mutually exclusive, differ in the extent to which they
Chapitre 6: Troisième article
119
stress the importance of particular types of factors in the development of burnout. As
suggested by the definition of burnout given above, the JD-R model approaches
burnout from an individual, interpersonal and organizational perspectives. These
perspectives emphasize the role of factors most relevant to management (e.g.,
employees’ perceptions, job features, team work, supervisory and organizational
practices).
The classic models of occupational stress, such as the demand-control model
of Karasek (1979), postulate an interaction effect between demands and resources. In
contrast, the JD-R model posits that burnout results from the independent additive
effect of two types of stress factors (Halbesleben & Buckley, 2004) related to work
and to the context in which the work is performed, namely excessive demands and
insufficient resources. Studies based on the JD-R model demonstrate that interaction
between
demands
and
resources
have
no
significant
effect,
or
a very weak effect at best, on burnout (cf. Bakker, Demerouti, Taris, Schaufeli, &
Schreurs, 2003; Bakker et al., 2004). second important postulate of the JD-R model
is that two distinct yet related processes contribute to the development of burnout (cf.
Schaufeli & Bakker, 2004). The energetic process originates from demands and is
mainly centered on emotional exhaustion. In contrast, the motivational process
originates from resources and is mainly centered on depersonalization. We postulate
that the two dimensions of burnout are linked indirectly to the intention to leave the
profession via psychosomatic complaints, associated with the energetic process, and
via professional commitment, associated with the motivational process.
Chapitre 6: Troisième article
120
6.2.2.1 Demands and resources
Contrary to most studies conducted on the JD-R model, our research seeks to
understand the particular influence of specific demands and resources on the two
burnout components rather than the influence of demands and resources as two
higher-order constructs. Our choice of specific demands and resources is based on the
theoretical work of Leiter and Maslach (2004), who proposed a structured approach
to organizational predictors of job burnout. Their model postulates that burnout
results from the misfit between the person and the job in six areas of worklife:
workload, control, reward, community, fairness, and values. Also, a review of the
literature has enabled us to identify a number of important stress factors that nurses
purportedly face on a regular basis (chronic stress factors). In this study, we consider
two types of demands and resources - 1) psychological aspects associated with work
and 2) aspects related to the social environment - which cover the six areas of
worklife mentioned above. Thus, as Figure 1 indicates, the two types of demands
included in our research model relate to the workload (quantitative overload), control
(role ambiguity and role conflict, which are combined to capture overall role stress),
community (hostility), and workload and values (work interference with family) areas
of Leiter and Maslach’s (2004) model. The two types of resources considered in our
study relate to the workload (sense of competence), control (decision-making
authority and impact), reward and fairness (recognition), community (support) and
values (meaning) areas of Leiter and Maslach’s (2004) model.
Chapitre 6: Troisième article
121
6.2.2.2 Energetic process
The influence of demands and resources on emotional exhaustion, which is
associated with psychosomatic complaints, is supported by the conservation of
resources (COR) theory (Hobfoll & Freedy, 1993). According to the COR theory,
individuals will invest the resources they own in order to protect resources, limit loss
of resources or gain resources. Mobilization of personal energies and other resources
to meet chronic demands induces a “loss spiral” phenomenon that affects reserves of
resources because they tend to co-travel in resource caravans (cf. Hobfoll, 2002).
When individuals are exposed to chronic demands, a process of depletion and
wearing down of their intrinsic energy resources develops over time. Because
personal energies, as secondary resources, determine the access and use of primary
resources such as social support (cf. Hobfoll & Shirom, 2000), a state of exhaustion
of personal energies on the emotional, physical and cognitive levels should coincide
with a state of impoverishment of other kinds of resources.
In support of this theoretical argument, the meta-analysis by Lee and Ashforth
(1996) demonstrates that demands are positively linked to emotional exhaustion (rc =
.21 to .65), whereas resources such as various types of support and intrinsic
motivational characteristics at work are negatively correlated (rc = -.10 to -.48) with
this dimension. Furthermore, in their attempt to validate the JD-R model using an
alternative measure of burnout applicable to any worker, Demerouti et al. (2001)
found that job demands are positively related to exhaustion. Although they report a
non-significant relationship between job resources and exhaustion, Bakker et al.
Chapitre 6: Troisième article
122
(2004) found in a later study that job resources had an important main effect, albeit
lesser than that of job demands, on exhaustion. Consequently, we formulate the
following hypotheses:
H1: Demands are positively associated with emotional exhaustion.
H2: Resources are negatively and less strongly associated with emotional
exhaustion than demands.
The control mode (cf. Hockey, 1997) that entails maximization of effort to
maintain the usual level of performance when individuals face demanding work
conditions can be adaptive over the short term. However, the habitual recourse to this
type of response gradually leads to emotional, physical and cognitive exhaustion that
incurs health costs. Accordingly, several studies report a positive relationship
between emotional exhaustion and psychosomatic complaints (e.g., De Jonge,
Janssen, & Van Breukelen, 1996; Landsbergis, 1988; Schaufeli & Van Dierendonck,
1993). Psychosomatic complaints refer to subjectively measured physical disorders
(e.g., backache, migraines, respiratory and digestive problems) that result from
frequent and/or prolonged psychophysiological arousal (Schaufeli et Enzmann,
1998). The following hypothesis is therefore formulated:
H3: Emotional exhaustion is positively linked to psychosomatic complaints.
Moreover, a causal link between emotional exhaustion and depersonalization
has been incorporated in the JD-R model (cf. Bakker et al., 2004), implying that the
two basic processes proposed by the model are not totally independent. Consistent
with
the
stress-strain-coping
paradigm
(cf.
Lazarus
&
Folkman,
1984),
depersonalization can be seen as a defensive adaptation strategy centered on emotions
Chapitre 6: Troisième article
123
intended to end a persistent state of emotional exhaustion (Cherniss, 1980; Leiter &
Maslach, 1988; Shirom, 2003). This coping strategy is likely to be used when
resources are exhausted (Cordes & Dougherty, 1993) and when the context does not
offer the possibility of adopting socially acceptable strategies (e.g., reducing work
hours, rotating positions) (Maslach, 1976). Whereas nurses require a sympathetic but
distant attitude to perform adequately in the health care sector (Lief & Fox, 1963),
depersonalization is a dysfunctional and relatively permanent attitude (Schaufeli &
Taris, 2005). It is characterized by feelings of insensitivity and impersonal responses
to the patients. According to Cherniss (1980), professionals adopt this attitude, even
if the aid relationship with the client is not the major source of stress, in an attempt to
block an additional emotional demand that would aggravate the state of exhaustion. A
study conducted by Eisenstat and Felner (1984) supports this proposition. Moreover,
the meta-analysis by Lee and Ashforth (1996) demonstrates that the emotional
exhaustion and depersonalization dimensions are positively linked (rc = .64) and the
temporal nature of their relationship is supported by longitudinal evidence (e.g.,
Taris, Le Blanc, Schaufeli, & Schreurs, 2005). We, therefore, formulate the following
hypothesis:
H4: Emotional exhaustion is positively associated with depersonalization.
6.2.2.3 Motivational process
Job resources may play either an extrinsic motivational role because they are
instrumental in achieving goals or an intrinsic motivational role because they fulfill
Chapitre 6: Troisième article
124
basic human needs. Two theories, the goal setting theory and the self-determination
theory, thus support the role of resources in this motivational process.
According to the goal setting theory (Locke & Latham, 1990), resources
increase extrinsic motivation at work by facilitating the attainment of objectives.
Thus, insufficient resources to face chronically imposed demands compromise the
attainment of a level of performance, which would give access to valued extrinsic
rewards. This situation is likely to create a feeling of failure and frustration that
should undermine employee’s extrinsic motivation. Demerouti et al. (2001) assert
that disengagement from work may represent an important self-protection mechanism
intended to prevent employee frustration over difficulty in attaining performance
objectives. Among nurses, as the objectives are directly linked to patient care, and
because it is particularly difficult to accomplish relational work with patients in the
current health care context (Darras et al., 2001-2003; White, 1993), disengagement is
likely to occur in this sphere of professional activity and therefore cause
depersonalization of contact with patients.
In contrast, the self-determination theory (Ryan & Deci, 2004) stipulates that
resources related to the execution of work that satisfy the needs for autonomy,
competence and interpersonal attachment favour personal growth and increase
intrinsic motivation at work. For example, the level of decision-making authority
given to nurses as part of their work with patients satisfies their need for autonomy,
gives them an opportunity for self-development and increases their motivation linked
to the accomplishment of the task. Inversely, the lack of resources related to these
basic human needs might encourage disengagement from a professional activity that
Chapitre 6: Troisième article
125
is not intrinsically satisfying and may give rise to an attitude of detachment toward
the patients. In a study of 180 nurses, Maslach and Jackson (1981) found that
satisfaction with personal growth opportunities was negatively correlated with
depersonalization. In addition, Blais et al., (1993) observed that three types of
intrinsic motivation (accomplishment, stimulation, learning) were negatively linked
to depersonalization.
On the basis of the previous theoretical arguments and given that the metaanalysis by Lee and Ashforth (1996) shows that resources such as various types of
support and intrinsic motivational characteristics of work are negatively linked to
depersonalization (rc = -.08, to -.46), we posit the following hypothesis:
H5: Resources are negatively associated with depersonalization.
If depersonalization is a protective adaptation strategy, it, nonetheless, causes
secondary effects on health. Nurses are then caught in a psychological conflict driven
by two factors. First, to avoid aggravating their state of emotional exhaustion, they
seek to anaesthetize all feelings of compassion for their patients. Secondly, nurses
who are attached to their role of care giver (White, 1993) and were trained in the
principles of holistic medicine (Lief & Fox, 1963) may struggle with the inability to
perform relational work with the patients. In support of these arguments,
depersonalization has been shown to be positively associated with psychosomatic
complaints (cf. Landsbergis, 1988; Schaufeli & Van Dierendonck, 1993). We,
therefore, formulate the following hypothesis:
H6:
Depersonalization
complaints.
is
positively
associated
with
psychosomatic
Chapitre 6: Troisième article
126
6.2.2.4 Professional commitment
Meyer and Herscovitch (2001) maintain that professional commitment is a
force - the psychological state of desire - that causes employees to want to maintain
their membership in a professional group. The authors argue that this state of desire is
explained by mechanisms such as involvement (intrinsic motivation), shared values
and identification. The links between resources and professional commitment via
depersonalization can be explained by low intrinsic motivation and unshared values.
The absence of resources that favour personal growth may decrease intrinsic
motivation and encourage disengagement from a professional activity that is not
satisfactory. Disengagement may first translate into depersonalization of contact with
the patients, which would initiate a process of erosion of professional commitment.
Further, a mechanism of unshared values may also be at play. Nurses having reached
the stage of depersonalization may struggle with their inability to perform the role of
care giver as they conceive it and become aware of a gap between the type of nurse
they aspire to be and that which they have become over time. In contrast, the links
between demands and professional commitment via emotional exhaustion can be
explained by a mechanism of identification with the profession. Indeed, the chronic
nature of demands and a persistent state of emotional exhaustion may lead nurses to
doubt that they are suited to this type of work. Under conditions of prolonged failure
of congruence with identity standard, identity theory predicts that a change in identity
standard will most likely occur (Burke & Reitzes, 1991). Low commitment to the
nurse identity should then produce weak ties to the nursing profession. Lastly, at the
Chapitre 6: Troisième article
127
empirical level, the meta-analysis by Lee, Carswell and Allen (2000) shows that
occupational (affective) commitment is negatively linked to depersonalization (rc = .37) and emotional exhaustion (rc = -.44). We, therefore, formulate the following
hypotheses:
H7: Depersonalization is negatively associated with professional commitment.
H8: Emotional exhaustion is negatively associated with professional
commitment.
6.2.2.5 Intention to leave the profession
Intentions can be viewed as statements about actions that one would take with
respect to an object (Fishbein, 1967). According to Fishbein (1967), behavioral
intentions represent the conative component of an individual’s attitude toward an
object. As such, they are a mode of mental functioning, like cognitions and affects
(cf. Snow & Farr, 1987), which are assumed to capture the individual’s motivation to
perform a behavior. In other words, behavioral intentions reflect a psychological state
of desire.
To our knowledge, only two studies - albeit ones with important limitations have investigated the link between burnout and intention to leave the nursing
profession. First, the qualitative study of Cheung (2004), involving semi-structured
interviews with 29 nurses, indicates that burnout may be an important factor in the
decision to leave the profession. Second, in their longitudinal study of 146 employees
in nursing care (including 93 nurses), Krausz, Koslowsky, Shalom and Elyakim
Chapitre 6: Troisième article
128
(1995) observed that emotional, physical and mental exhaustion at time 1 was
positively correlated with the intention o leave the profession one year later, at time 2
(.22). Unfortunately, they adopted a one-dimensional definition of burnout, excluding
the depersonalization component, which has been vigorously debated (cf. Green et
al., 1991; Maslach et al., 2001; Schaufeli & Taris, 2005). More importantly, they
posited a direct link between burnout and intention to leave the profession. Outside
the nursing profession, Jackson, Schwab and Schuler (1986) have found that
emotional exhaustion and depersonalization at time 1 significantly predicted at time 2
elementary and secondary school teachers’ thoughts about leaving education.
There is a theoretical reason to believe that the relationship between burnout
and the intention to leave the profession is not direct, and that it can be explained by
attitudinal change toward the profession. In Rhodes and Doering’s (1983) model, the
intention to change careers is directly influenced by career satisfaction. However, a
validation study of their model performed on teachers demonstrates that the
relationship is weak (β = -.14, p .05) (cf. Rhodes & Doering, 1993). n contrast,
although it examines correlates, the meta-analysis by Lee et al. (2000) shows that
occupational commitment is a fairly robust predictor of intention to leave the
profession (rc = -.62). Moreover, if the career change model of Rhodes and Doering
(1983) takes into account several personal factors (e.g., seniority, work-family
conflict), it does not consider health problems. Yet, such problems may lead nurses to
leave the profession (cf. Cheung, 2004; Fimian, Fasteneau, & Thomas, 1988). deed,
nurses may conclude that the only miracle cure for their persistent health problems
Chapitre 6: Troisième article
129
would be to grudgingly abandon their profession. Based on these arguments and
evidence, we formulate the following hypotheses:
H9: Professional commitment is negatively associated with the intention
leave the profession.
H10: Psychosomatic complaints are positively associated with the intention
leave the profession.
6.3
Methodology
6.3.1 Participants
The study was conducted on a random sample of unionized registered nurses
(RNs) working in the Canadian public health care sector, stratified by mission and
size of the institution to ensure representativeness. Of the 6,546 nurses solicited with
the union’s help, 2,175 returned a completed questionnaire, for a response rate of
33.2%, which is satisfactory considering the length of the questionnaire. In addition,
the likelihood of getting nurses, “(…) so sick and tired of being studied to death (…)”
and not seeing any results (Kerr et al., 2002, p.16), to fill in another detailed survey
was very low. The limited information available indicates that respondents do not
differ from the overall population in terms of gender and age. Given that health care
institutions present different contexts depending on their mission, only nurses
working in hospitals and having direct contact with patients were retained for this
Chapitre 6: Troisième article
130
study, for a final sample of 1,636 respondents with the following characteristics 8:
92% are female, the average age is 41, 74% are married or live with a common law
partner, 56% have college training, 56% have full time job status and 46% work on a
day shift. They have on average 8 years of seniority in their current position, 16 years
of seniority at their institution and 18 years of experience as a nursing professional.
6.3.2 Measures
Unless otherwise specified, nurses were asked to express their level of
agreement with each statement on a continuum ranging from Strongly disagree (1) to
Strongly agree (7).
Demands. Comprising three items, the quantitative overload scale was
developed based on the instrument designed by Caplan et al., (1980) (e.g., “I
regularly feel overloaded by my work”). Role stress engendered by the ongoing
reform of the Canadian public health care system is measured using four items
developed for this study, inspired by the work of Collerette and Schneider (1996) and
from instruments designed by Rizzo, House, and Lirtzman (1970) to measure role
ambiguity and role conflict. The scale includes items such as “The role of everyone in
the new organizational functioning is clear for all the employees” and “I do not know
how to realize the proposed changes; I do not have the necessary materials or
resources.” The three items that form the work interference with family scale are
taken from the instrument designed by Gutek, Searle and Klepa (1991) and enhanced
8
Percentages reported are calculated based on valid cases.
Chapitre 6: Troisième article
131
by Carlson and Perrewé (1999) (e.g., “My job takes time that I would like to spend
with my family”). Scales linked to hostility from physicians and patients each
includes three items taken from the Inventory of Stressful Events developed by
Motowidlo et al., (1986) to measure the frequency of stressful events experienced by
nursing staff at hospitals. The scales capture different forms of aggression toward
nurses ranging in severity and include revised items such as “Some doctors publicly
criticize my nursing care” and “Some patients use abusive language or inappropriate
gestures with me”. Nurses were instructed to indicate the frequency of specific
behaviors during the last twelve months, on a continuum ranging from Never (1) to A
few times a month (4) to Every day (7).
Resources. Four resources, each measured with three items, are components
of the psychological empowerment (PE) construct (cf. Thomas & Velthouse, 1990).
Statements measuring competence (e.g., “I have mastered the skills necessary for my
job”) and meaning (e.g., “The work I do is very important to me) are taken from the
instrument developed by Spreitzer (1995). For the other dimensions of PE, we refer
to the less restrictive concept of decision-making authority that designates the degree
of influence both on the way the work must be done and on its nature (Theorell,
2003). One such item is “I have enough authority to fulfil the responsibilities
assigned to me”. The fourth dimension, impact, is defined as the employees’ feeling
that the accomplishment of their personal work has significant repercussions on their
environment, in order to be consistent with Thomas and Velthouse’s concept of PE,
which consists in four cognitions associated with the task. A sample item is “The
fulfilment of my work tasks enables me to have considerable impact on my work
Chapitre 6: Troisième article
132
environment”. Support from supervisor and support from colleagues are each
measured using three items taken from the instrument developed by Eisenberger et
al., (1986) to measure perceived organizational support, adapted to each of the
support sources. The items selected reflect emotional support (e.g., “My supervisor
really cares about my well-being”), esteem support (e.g., “My colleagues care about
my opinion”) and instrumental support (e.g., “If I have a problem, I can get help from
my supervisor”). Measurement scales linked to recognition by physicians and
patients, created for this study, each includes three items (e.g., “The doctors usually
give me credit for my accomplishments”; “The patients often show their satisfaction
with the services that I provide to them”).
Endogenous variables. Items linked to the two dimensions of burnout are
taken from the Maslach Burnout Inventory - Human Services Survey (MBI-HSS)
(Maslach & Jackson, 1996). Thus, emotional exhaustion (e.g., “I feel emotionally
drained from my work”) and depersonalization (e.g., “I feel I treat some recipients as
if they were impersonal objects”) are each measured using the four items that present
the greatest factor loadings. Nurses were asked to indicate the frequency of symptoms
experienced in the last twelve months, on a continuum ranging from Never (1) to A
few times a month (4) to Each day (7). Professional commitment is measured using
four items taken from the instrument of Meyer et al. (1993), signed to measure
affective commitment toward the occupation and selected based on their factor
loading (e.g., “I am proud to be in the nursing profession”). On the basis of the
psychosomatic complaints included in the Depression Inventory of Beck (1978), three
items were formulated and measured on a frequency scale rather than by the severity
Chapitre 6: Troisième article
133
of the symptom. The items refer to sleep, appetite and health disturbances (backache,
migraines, respiratory and digestive problems). Nurses were asked to indicate the
frequency of symptoms experienced in the last twelve months, on a continuum
ranging from Never (1) to A few times a month (4) to Each day (7). Lastly, the three
items that refer to the intention leave the profession were developed by Meyer et al.
(1993) and adapted to an agreement-disagreement scale (e.g., “It is possible that I quit
the nursing profession within the next year”).
Control variables. We measure the influence of socio-demographic data and
characteristics related to the job on the five endogenous variables included in our
model. These control variables are gender, age, marital status, number of dependents
and children under age 12, level of education, job status, work schedule, seniority in
the current position and at the institution, work experience as a nurse, supervision
responsibilities, clinical field, hospital and administrative region.
6.3.3 Analyses
Structural equation modeling (SEM) was used as a statistical methodology. As
a multivariate method, SEM provides a way to conduct a simultaneous analysis of the
entire system of variables. However, in contrast to traditional multivariate techniques,
the SEM method explicitly takes measurement error into account when statistically
analyzing data and incorporates both unobserved (i.e. latent) and observed variables.
A preliminary phase of development of the model was performed based on a
calibration sample (N = 818) that we formed by randomly dividing the sample of
Chapitre 6: Troisième article
134
1,636 nurses into two. Consistent with the approach suggested by Anderson and
Gerbing (1988), we assessed the quality of fit of the measurement model using
confirmatory factor analysis, then estimated the structural model. In a second strictly
confirmatory phase, we tested the final structural model obtained in phase 1 on the
second portion of our sample (N = 818) (cf. Bagozzi & Yi, 1988) and verified our
hypotheses. The quality of fit of the models was evaluated based on several indices.
Accordingly, a Root Mean Square Error of Approximation (RMSEA) below .05 is a
sign of a good degree of fit, while values of up to .08 indicate a reasonable error of
approximation in the population (Byrne, 1998). A Normed Fit Index (NFI) and a
Comparative Fit Index (CFI) situated between .90 and 1 also indicate the presence of
a well-fitted model (Bentler, 1992; Hu & Bentler, 1995). We also report the classic
chi-square (χ2) statistic to compare the quality of fit of nested models.
6.4
Results
6.4.1 Preliminary analysis
Because results obtained with self-reported measures might be confounded by
common method variance, Harman’s one-factor test was performed (cf. Podsakoff &
Organ, 1986) on the 54 scale items used to assess the 18 variables considered in the
present study. Using an eigenvalue greater than the 1.00 cut-off criterion, 15 factors
could be justified across the two subsamples. Although Podsakoff and Organ argue
that the results of such a test cannot be interpreted unequivocally, they do,
Chapitre 6: Troisième article
135
nonetheless, provide some evidence that common method variance does not account
for most of the interrelationships between the variables since the items used to assess
the independent, mediating and dependent variables do not tend to load on a single
general factor.
6.4.2 Development Phase of the Model
6.4.2.1 Measurement model
The quality of fit of the measurement model was evaluated using first-order
confirmatory factor analysis (CFA). First, we conducted separate analyses for each of
the 18 latent variables included in our research model. For each of the scales
measuring role stress, depersonalization and professional commitment, one item was
eliminated because of a lack of fit of the model, low variance explained by the
construct or poor factor loading. We then conducted a global analysis by including all
the 18 variables. In this 18-factor model, each item was associated with the factor
defined a priori, and factor loadings ranged from .53 to .93 (p .05), thus meeting the
minimum threshold of .50 recommended by Roussel et al. (2002). he level of
reliability of internal consistency of the latent variables (Jöreskog’s rho) varied
between .71 and .91, which satisfies the minimum threshold of .70 recommended by
Roussel et al. (2002), except for psychosomatic complaints (.66). Regarding the
quality of fit, the indices suggested a satisfactory level of fit [ 2(1224) = 2480.76,
p .001; RMSEA = .036; NFI = .96; CFI = .98].
Chapitre 6: Troisième article
136
The last step of the measurement model analysis was to compare our proposed
18-factor model with simpler models. Because theory justifies the presence of two
meta-constructs, burnout and psychological empowerment (PE), we assessed the fit
of 1) a 17-factor model, where items related to the two components of burnout are
specified to load on the same factor; and of 2) a 15-factor model, where items related
to the four dimensions of PE are specified to load on the same factor. The 17-factor
model provided a less satisfactory fit [ 2(1225) = 2673.68, p .001; RMSEA = .038;
NFI = .96; CFI = .98] than the proposed 18-factor model (Δ 2(1) = 192.92, p .001).
The results obtained for the 15-factor model [ 2(1230) = 4781.15, p .001; RMSEA =
.064; NFI = .93; CFI = .94] indicated an even greater deterioration in the quality of fit
(Δ 2(6) = 2300.39, p .001). Also, because a strong correlation exists between
emotional exhaustion and psychosomatic complaints ( = .81, p .001), we evaluated
the fit of another 17-factor model, where items related to these two variables are
specified to load on the same factor. Likewise, this second 17-factor model showed a
less satisfactory fit (
2
(1225)=2555.14,
than the proposed 18-factor model (
p .001; RMSEA=.037 ; NFI=.96 ; CFI=.98)
2
(1)=74.38,
p .001). Therefore, the 18-factor
model was used to test our hypotheses.
6.4.2.2 Structural model
Before reporting results, two comments need to be made regarding the
procedure followed for the structural model analysis. First, during the test of the
structural model on the calibration sample, given that the inclusion of the control
Chapitre 6: Troisième article
137
variables did not induce substantial change, either in terms of the quality of fit of the
model or in the value and significance of estimated parameters, we decided to
exclude the control variables from the structural model for parsimony purposes.
Secondly, the solution obtained during the first test of the structural model proved to
be inadmissible due to problems of estimation of the structural parameters linking the
two dimensions of burnout to professional commitment (very high value, suppressor
effect), which can occur in the case of overly high co-linearity between the variables
of the model (Bentler & Chou, 1987). As recommended by Roussel et al. (2002), the
use of a single composite indicator (mean of the scores of each item) for
depersonalization and emotional exhaustion solved this problem. For these latent
variables, the factor loading was set at equal to the square root of the alpha coefficient
of reliability (λy = α) and the variance of the measurement error was set at equal to
the value obtained using the following formula: θε = (1-α)(SD2) (cf. Netemeyer,
Johnston, & Burton, 1990).
The quality of fit of the structural model to the data of the calibration sample
(N=818) appears satisfactory [ 2(1143) = 3799.41, p .001; RMSEA = .062; NFI =
.93; CFI = .95]. However, modification indices concerning the structural relations of
the model revealed an under-estimated relationship, that between meaning and
professional commitment (MI = 109.68). From a theoretical standpoint, this direct
relationship is plausible because a nurse who considers her work important will
obviously be committed to her profession. Consequently, we freed this link. The
results obtained for the revised model [ 2(1142) = 3651.82, p .001; RMSEA = .060;
NFI = .93; CFI = .95] indicate a significant improvement in the quality of fit of the
Chapitre 6: Troisième article
138
model (Δ 2(1) = 147.59, p .001). In addition, there were no lingering major
modification indices.
6.4.3 Confirmatory Phase
6.4.3.1 Structural model
Table I (p.139) presents the descriptive statistics and correlations between
variables of the structural model for the validation sample (N=818). The quality of fit
of the revised structural model to the data of the validation sample appears
satisfactory [ 2(1142) = 3783.79, p .001; RMSEA = .061; NFI = .93; CFI = .95] and
the variance explained for each of the endogenous variables is: 35% for
depersonalization, 48% for emotional exhaustion, 39% for professional commitment,
69% for psychosomatic complaints and 48% for intention to leave the profession. The
research hypotheses are discussed based on the results obtained for the different
structural parameters estimated (see Figure 2, p.140).
Table I
Descriptive statistics and correlations (validation sample)
Variables
1. COMP
2. AUTH
3. MEAN
4. IMP
5. SUPSU
6. SUPCO
7. RECPH
8. RECPT
9. QTOVL
10. ROLST
11. INTFA
12. HOSPH
13. HOSPT
14. DEPER
15. ÉMOEX
16. PRCOM
17. PSYCO
18. INTLV
M
SD
5.92
.88
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
.45**
.50**
.47**
.13**
.24**
.24**
.19**
-.11**
-.20**
-.14**
-.15**
-.09*
-.17**
-.14**
.24**
-.14**
-.16**
5.24
1.20
.53**
6.00
.98
.58**
.55**
.47**
.46**
.31**
.36**
.23**
.20**
-.26**
-.36**
-.22**
-.20**
-.12**
-.21**
-.26**
.27**
-.21**
-.24**
.61**
.26**
.33**
.20**
.34**
-.14**
-.26**
-.15**
-.10**
-.17**
-.36**
-.24**
.55**
-.22**
5.62
1.04
.53**
.54**
.74**
-.35**
.28**
.28**
.19**
.27**
-.10**
-.24**
-.11**
-.05
-.11**
-.23**
-.13**
.33**
-.11**
4.51
1.59
.16**
.36**
.31**
.35**
-.20**
.38**
.23**
.11**
-.24**
-.43**
-.18**
-.14**
-.04
-.23**
-.25**
.25**
-.17**
5.56
1.09
.28**
.41**
.38**
.35**
.44**
-.22**
.26**
.27**
-.17**
-.29**
-.18**
-.18**
-.09*
-.18**
-.22**
.32**
-.20**
4.26
1.55
-.19**
.26**
.27**
.22**
.23**
.26**
.28**
5.99
.96
.23**
.25**
.41**
.33**
.15**
.33**
.20**
-.21**
-.23**
-.22**
-.36**
-.15**
-.28**
-.25**
.21**
-.18**
-.24**
-.07
-.19**
-.05
-.14**
-.29**
-.30**
-.12**
.31**
-.16**
4.71
1.39
-.14**
-.30**
-.16**
-.12**
-.27**
-.18**
.47**
.22**
.12**
.22**
.60**
-.15**
.31**
4.40
1.62
-.23**
-.42**
-.32**
-.34**
.26**
.23**
.16**
.11**
.25**
.33**
-.24**
.22**
1.51
.51
-.16**
-.24**
-.16**
.21**
.18**
.08*
.21**
.56**
-.19**
.35**
2.07
.86
.29**
-.24**
-.30**
4.07
1.25
-.12**
-.17**
.21**
.24**
.26**
-.09**
.21**
.16**
.27**
.19**
-.10**
.24**
1.18
.18
-.22**
.15**
.36**
-.33**
.33**
3.79
1.33
.32**
-.31**
.62**
5.99
.41**
-.28**
-.47**
-.17**
.21**
-.24**
-.07
-.51**
-.33**
-.25**
-.21**
.36**
-.12**
-.18**
-.20**
-.21**
-.05
.51**
.24**
-.15**
-.14**
-.17**
-.22**
-.40**
-.21**
.27**
.19**
.19**
-.22**
-.15**
-.06
-.13**
-.21**
-.40**
.17**
.14**
.12**
-.31**
-.47**
-.26**
-.29**
-.26**
-.35**
-.33**
.34**
.29**
-.16**
-.30**
-.29**
-.16**
-.28**
-.25**
-.27**
-.14**
.72**
1.16
.27**
.30**
.63**
.41**
.30**
.37**
.25**
.37**
2.49
1.01
-.21**
-.34**
-.32**
-.19**
-.25**
-.30**
-.24**
-.21**
2.04
1.07
-.19**
-.30**
-.45**
-.27**
-.28**
-.21**
-.26**
-.23**
.33**
.31**
.38**
.35**
.41**
.35**
.54**
.31**
.24**
.53**
-.18**
-.29**
-.19**
-.13**
-.14**
-.45**
-.37**
.45**
.29**
.43**
.37**
.37**
.60**
.81**
-.37**
.31**
.23**
.19**
.53**
.53**
-.62**
.35**
.30**
.33**
.50**
Note. N = 818. Pearson’s correlations (r) are reported above the diagonal. Correlations between latent variables ( ) are reported below the diagonal. COMP = competence;
AUTH = decision-making authority; MEAN = meaning; IMP = impact; SUPSU = support from supervisor; SUPCO = support from colleagues; RECPH = recognition by
physicians; RECPT = recognition by patients; QTOVL = quantitative overload; ROLST = role stress; INTFA = work interference with family; HOSPH = hostility from
physicians; HOSPT = hostility from patients; DEPER = depersonalization; EMOEX = emotional exhaustion; PRCOM = professional commitment; PSYCO =
psychosomatic complaints; INTLV = intention to leave the profession.
* p .05. ** p .01.
Chapitre 6: Troisième article
140
Figure 29
Results of the test of the structural model (validation sample)
Competence
.51***
Decision-making authority
-.43***
RESOURCES
Meaning
-.16***
.15*/ -.16**
Impact
Support supervisor
Professional
commitment
Depersonalization
-.08*
-.20***
-.50***
-.17***
Support colleagues
-.13**
Recognition physicians
-26***
.36***
Intention to
leave the
profession
Recognition patients
.14* / -.03
.22***
Quantitative overload
.57***
.37***
DEMANDS
Role stress
Work interference
with family
Hostility physicians
.08*
.71***
.30***
.12***
Emotional
exhaustion
Psychosomatic
complaints
.11**
Hostility patients
Note. Significant structural relationships between exogenous and endogenous variables (validation
sample, N = 818). Parameters are reported in standardized form. Path coefficients in bold from impact
to the two components of burnout: coefficients obtained when significant relationships between
meaning and the endogenous variables are omitted from the analysis. *p .05. ** p .01. *** p .00.
9
For clarity reason, only the significant structural relationships are illustrated. The non-significant
relationships are: 1) with depersonalization: competence (γ = .05), decision-making authority (γ = .01),
support from colleagues (γ = .05); and 2) with emotional exhaustion: competence (γ = .07), decisionmaking authority (γ = -.02), support from the supervisor (γ = -.04), support from colleagues (γ = -.05),
recognition by physicians (γ = -.04), recognition by patients (γ = .02).
Chapitre 6: Troisième article
Energetic process originating from demands. Hypothesis 1, which posits that
demands are positively associated with emotional exhaustion, is confirmed.
Quantitative overload, role stress, work interference with family, hostility from
physicians and patients are positively linked to emotional exhaustion. Regarding
Hypotheses 3 and 4, emotional exhaustion is positively linked to psychosomatic
complaints and depersonalization, which confirms these two hypotheses. The results
also confirm Hypothesis 8, in that emotional exhaustion is negatively associated with
professional commitment. Hypothesis 10, which stipulates that psychosomatic
complaints are positively associated with the intention to leave the profession, is
confirmed as well.
Motivational process originating from resources. Hypothesis 2, which
stipulates that resources are negatively and less strongly associated with emotional
exhaustion than demands, is not supported. Of the eight resources examined, only the
meaning of work is negatively associated with emotional exhaustion and the
relationship is stronger than three out of five demands, namely role stress and
hostility from physicians and patients. Moreover, we observed a positive but weak
relationship (γ = .14, p .05) between impact and emotional exhaustion. Nonetheless,
the correlation between these two variables is negative ( = -.16, p .001), which is
generally considered a sign of a suppressor effect that may be attributable to
multicollinearity (Bentler & Chou, 1987). Indeed, supplementary analyses revealed
that the problem originates from the strong correlation between meaning and impact
( = .74, p .001). When significant relations between meaning and the endogenous
variables are omitted, the relationship between impact and emotional exhaustion
141
Chapitre 6: Troisième article
142
becomes negative and non-significant (γ = -.03, ns), a result that is not consistent
with our hypothesis.
Hypothesis 5, which posits that resources are negatively associated with
depersonalization, is partially supported. In fact, four out of eight resources examined
are negatively related to depersonalization: meaning of work, support from supervisor
and recognition by physicians and patients. Results concerning competence, decisionmaking authority and support from colleagues are not consistent with the hypothesis.
Here again, we observed a suppressor effect in that the relationship between impact
and depersonalization is positive but weak (γ = .15, p .05), whereas the correlation is
negative ( = -.26, p .001). When significant relations between meaning and the
endogenous
variables
are
omitted,
the
relationship
between
impact
and
depersonalization becomes negative and remains significant (γ = -.16, p .01), a result
in line with our hypothesis.
Further, Hypothesis 6 is confirmed since depersonalization is positively
associated with psychosomatic complaints. The results also support Hypothesis 7,
whereas depersonalization is negatively associated with professional commitment.
Hypothesis 9, which states that professional commitment is negatively associated
with the intention to leave the profession, is confirmed as well. It is worth nothing
that the positive relationship between meaning and professional commitment,
observed a posteriori, is one of the strongest relations in our model.
Chapitre 6: Troisième article
6.4.3.2 Additional analyses
Complementary tests of the mediation effects implied by our model
demonstrated that emotional exhaustion has a greater total effect on intention to leave
the profession (.41, p<.001) than depersonalization (.28, p<.001). Further, we
examined the robustness of the ultimate structural model, including only the
significant relations, by applying a constraint of equality of the structural parameters
of the calibration and validation samples. The results of the multigroup analysis
reveal that imposition of a constraint of invariance of the structural parameters does
not cause significant deterioration in the quality of fit in comparison with the model
where the parameters are free to vary (Δ 2(20) = 25.87, p .001). This test therefore
demonstrates that the structural relations of the ultimate structural model are stable
across the two groups, which reinforces our confidence in the validity of the model.
6.5
Discussion
In our view, this is the first quantitative study to establish links between stress
factors related to burnout, conceptualized as a syndrome of emotional exhaustion and
depersonalization, and nurses’ intention to leave their profession. By applying the job
demands-resources (JD-R) model of burnout (Demerouti et al., 2001), this study has
attempted to clarify the role of stress factors related to nurses’ work and social
environment in the development of burnout that have rarely been empirically
evaluated (i.e. psychological empowerment, recognition and hostility from physicians
143
Chapitre 6: Troisième article
and patients). Unlike most studies conducted on the JD-R model, our research has
examined the unique influence of specific demands and resources on the two burnout
components rather than the influence of demands and resources as two higher-order
constructs. Further, this study has attempted to shed light on the nature of the
relationship between burnout and intention to leave the profession by examining
intermediate linkages, which had previously not been done, to our knowledge.
6.5.1 Interpretation of results
Regarding the role of stress factors in the development of burnout, the results
are consistent with the proposition of Eisenstat and Felner (1983): demands indirectly
favour depersonalization via emotional exhaustion, and resources are associated with
a more humane treatment of patients, while evidence of direct links with emotional
exhaustion is sparse.
Stress factors as determinants of emotional exhaustion and, indirectly,
depersonalization. The five demands considered are positively linked to emotional
exhaustion, the relationship with quantitative overload being the strongest of all, and
emotional exhaustion is in turn positively associated with depersonalization.
However, the links between resources and emotional exhaustion are less robust and
warrant further discussion.
First, the meaning of work is the only resource that is negatively associated
with emotional exhaustion. The paucity of evidence of direct links between resources
144
Chapitre 6: Troisième article
145
and emotional exhaustion seems to argue against the concept of “loss spirals”
proposed by the conservation of resources theory. This theory suggests that a state of
exhaustion of personal energies should coincide with a state of impoverishment of
other kinds of resources. Nonetheless, it also states that because individuals strive to
protect themselves from resource loss, workers are more sensitive to workplace
phenomena that translate into losses for them than those that result in gains. Demands
threaten or cause a loss of resources that employees consider important, whereas
insufficient resources reduce only the possibilities of increasing their pool of existing
resources. The results indicate that this greater reactivity to demands than to
resources is true in the case of emotional exhaustion, as we note that on average
demands are more strongly linked to emotional exhaustion (
= .43) than are
resources ( = -.23). Nevertheless, in a context of quantitative overload, nurses seem
to be very sensitive to the absence of one resource in particular, the meaning of work,
to which their level of affective commitment to the profession is strongly linked.
Job resources as determinants of depersonalization. Four out of eight
resources are associated with depersonalization. Three of these resources are linked to
the psychological states critical to intrinsic motivation, postulated by the job
characteristics model (Hackman & Oldham, 1980): meaning, recognition and impact.
This model proposes that employees are intrinsically rewarded when they are aware
of their success in a task that has meaning or value to them. In our study, it is indeed
the meaning of work and recognition by physicians and patients that are significantly
linked to depersonalization. The impact of the task, whose significant effect on
depersonalization is negative when meaning is excluded from the model, also seems
Chapitre 6: Troisième article
to be a relevant motivational factor that would lead, according to the JCM, to
experienced meaningfulness of the work. Finally, support from the supervisor has a
lesser influence on depersonalization, barely exceeding that of colleagues, whose
effect is non-significant.
Nonetheless, contrary to our hypothesis, competence, decision-making
authority and support from colleagues are not linked to depersonalization.
Specifically, range restriction might account for the non-significant role of
competence in explaining both burnout components. Also, as Gagné, Senécal and
Koestner (1997) suggest, it is possible that the feeling of competence is a less
powerful motivational factor than being recognized as competent. In our study,
recognition by physicians and patients, rather than competence, is significantly linked
to depersonalization. The same type of argument can explain the fact that unlike
impact, decision-making authority has no significant effect on depersonalization. The
capacity to make decisions concerning their professional practice may be a
motivational factor for nurses in that the decisions they make have an important
impact on their work environment. The model of Liden and Arad (1996), which
focuses on the acquisition process of individual and collective power, supports this
argument. Their model positions choice, predicted by competence, as a direct
antecedent of impact: choice indicates potential power whereas impact reflects actual
power. Lastly, support from colleagues does not have a significant effect in the
model, unlike support from the supervisor. Beehr (1985) suggests that instrumental
support is likely to originate more from the supervisor, whereas emotional support is
mainly provided by colleagues. Thus, the role of the supervisor might be more critical
146
Chapitre 6: Troisième article
because supervisors possess more power to act on stress factors and improve
performance at work.
As a general observation, we note that the hypotheses pertaining to the
motivational process originating from job resources and mainly centered on
depersonalization did not receive full support while, in contrast, those pertaining to
the energetic process originating from job demands and mainly centered on emotional
exhaustion were all confirmed. This finding might be due to the fact that the items
used to measure the depersonalization dimension of the burnout syndrome are very
specific in that they make reference to the recipients of health care services. As a
dysfunctional attitude characterized by feelings of insensitivity and impersonal
responses to the patients, depersonalization might be more strongly prompted by a
lack of specific job resources such as the meaning of work, the recognition by
patients and the impact of the task.
Nature of the relation between burnout and intention to leave the profession.
The results indicate that emotional exhaustion and depersonalization are linked to
psychosomatic complaints and professional commitment, which are in turn associated
with the intention to leave the profession. The analysis of the strength of the path
coefficient obtained for two of the cross-links between the variables associated with
the energetic and motivational processes is quite illuminating. The results show that
emotional exhaustion has a much greater influence on psychosomatic complaints than
depersonalization, thus suggesting that this consequence of burnout is mainly
associated with the energetic process, accordingly to our research model. In contrast,
depersonalization is a slightly better predictor of professional commitment than
147
Chapitre 6: Troisième article
emotional exhaustion, which suggests that this consequence of burnout is mainly
associated with the motivational process, accordingly to our research model. Lastly,
we note that even though the intention to leave the profession is more strongly
predicted by professional commitment than psychosomatic complaints, the energetic
process involved in the development of burnout, nonetheless, plays a very important
role in the process that leads nurses to leave their profession. Indeed, results show
that of the two symptoms associated with burnout, emotional exhaustion has a greater
total influence on intention to leave the profession, although this influence is slightly
weaker than the direct effect of professional commitment on intention to leave.
Contextualization of results. Our study was conducted on a sample of
registered nurses belonging to the greatest regional union (who represents 84% of the
population targeted in this study) and working in hospitals belonging to the public
health care network. In Canada, it is the public sector, highly unionized, that hires the
majority of the nurses (87.8%) (Canadian Institute for Health Information, 2006).
Institutions belonging to this network markedly penetrated by one major union are
likely to resemble each other in terms of health care and human resources
management policies and practices. Hence, the burnout victims in our sample of
nurses were undoubtedly more inclined to consider leaving their profession instead of
moving from one health care institution to another. In fact, this is one important
reason why we choose to focus on professional commitment and intention to leave
the nursing profession as opposed to organizational commitment and turnover. In
contrast, in contexts where the private health care network predominates, it is quite
probable that nurses who experience this syndrome would first try to seek
148
Chapitre 6: Troisième article
intermediate solutions such as working for a healthcare facility that offers more
favourable task and work environments.
6.5.2 Study limitations
This study has some limitations that should be noted. First, response rate was
low, and we had access to limited information to assess the representativeness of our
sample. Second, the use of abbreviated scales for some constructs might have
affected our results although the latent variables showed a satisfactory level of
reliability of internal consistency, except for psychosomatic complaints. This last
variable was also highly correlated with feelings of being emotionally drained and
exhausted (either physically or cognitively) by one’s work. Such a strong relationship
would be expected given that psychosomatic complaints are caused at least partly by
mismanaged emotions. Nevertheless, the fact that the relationships between the
different variables, based on a priori set hypotheses, were generally confirmed
strengthen our confidence in regard to construct validity. Third, we used crosssectional data to examine presumed causal relationships. A longitudinal study would
have allowed us to measure actual withdrawal from the nursing profession and thus
better grasp the magnitude of the impact of current working conditions in the health
care system.
Moreover, two other important methodological aspects deserve discussion.
First, it may be argued that our results obtained with self-reported measures from a
single source might be confounded by common method variance. While we justify
149
Chapitre 6: Troisième article
the use of self-reported measures in our study by the phenomenological definition of
stress proposed by Lazarus and Folkman (1984) and the psychological nature of most
constructs included in our research model, respondents were asked to express their
judgment on relatively concrete attitudes, behaviors or facts, thus leaving little room
for inference (cf. Doty & Glick, 1998). Also, we provided some evidence, based on
the results of Harman’s one-factor test, that this bias might not pose a problem in the
present study. Second, the omission of negative affectivity in our research model may
represent another limitation, because this personality trait purportedly explains the
magnitude of relations observed between the perception of stress factors and selfreported symptoms (Burke et al., 1993; Watson et al., 1987). However, a number of
more recent studies have demonstrated that this affective disposition does not
significantly bias relations between self-reported measures of stress factors and
associated symptoms, to the extent that they are invalidated (e.g., De Jonge et al.,
2001; Schonfeld, 1996; Williams & Anderson, 1994; Williams et al., 1996).
6.5.3 Practical implications
The results obtained seem to suggest that excessive demands would be more
detrimental to individual well-being than insufficient resources would. Considerable
demands lead to emotional exhaustion and indirectly to depersonalization, whereas
the lack of resources has an influence mainly on depersonalization. Nonetheless, this
influence is much greater than that of emotional exhaustion. The percentage of
variance of depersonalization explained by resources and emotional exhaustion is
35%, whereas resources alone explain 28%. These results support an important
150
Chapitre 6: Troisième article
151
postulate of the JD-R model, namely that burnout results from the independent and
additive effect of two types of stress factors: demands are the most important
determinants
of
emotional
exhaustion,
whereas
resources
mainly
predict
depersonalization. As intention to leave the profession is more strongly predicted by
professional commitment than psychosomatic complaints, it is even more important
to increase resources that reinforce commitment toward the profession while
simultaneously preventing the development of a detached attitude toward the patients.
One may think that the reinforcement of professional commitment might be a
potential harmful suggestion. Yet, we argue that a strong sense of commitment to the
nursing profession that is mainly a response to external resources allocation will not
have the same negative impact on health than an exhaustive work-related personal
coping style such as over-commitment has been found to have (cf. Preckel et al.,
2007). The effort-reward-imbalance model (Siegrist, 1996) assumes that employees
characterized by over-commitment underestimate the external demands and
overestimate their own coping resources, thus contributing to the risk of prolonged
exposure to non-reciprocal exchange.
Moreover, of the two symptoms associated with burnout, emotional
exhaustion has a greater total influence on intention to leave the profession. Although
this influence is slightly weaker than the direct effect of professional commitment on
intention to leave, acting on demands seems to be equally important if the objective is
to retain nurses within the profession. Thus, a dual strategy is recommended: a
decrease in demands related to work and to the context in which the work takes place,
coupled with an increase in available resources. In particular, nurses’ tasks and role
Chapitre 6: Troisième article
should be restructured to reduce work overload and increase the meaning of their
work. The latter psychological resource appears to be crucial given its relationship
with the two burnout dimensions and professional commitment.
6.5.4 Research avenues
Regarding the development of burnout, some variables postulated but not
measured are worth integrating in subsequent validation studies of the JD-R model. A
more precise representation of the motivational process would entail measuring the
impact of resources on intrinsic and extrinsic motivation. Further, measuring the
greater effort deployed at work (cf. Büssing & Glaser, 2000) in response to chronic
demands could better capture the energetic process. As well, it is worth investigating
the types of primary resources that are likely to be sought and gained by mobilizing
additional emotional, physical and cognitive energy. Incidentally, an interesting
question concerns the dynamics between resources and demands. Specifically, how
are resources or demands interrelated within a single set? Could some combinations
of insufficient resources or excessive demands lead to a higher risk of burnout than
others? Another aspect of the JD-R model that needs to be explored in greater depth
is the role of resources in the development of burnout. It is possible that the resources
whose effect is non-significant in our research model may indirectly affect the two
components of burnout via their impact on demands. Although the effect of resources
on demands was not the subject of this research, one of the presumed functions of
resources is indeed to reduce demands (Bakker et al., 2004). Investigating these links
in a future study, supported by theoretical analysis, would be worthwhile.
152
CHAPITRE 7: DISCUSSION GÉNÉRALE
Les causes possibles d'épuisement professionnel ont été considérablement
investiguées depuis la “découverte” du syndrome au début des années 70. Pour
prédire l'épuisement professionnel, plusieurs études ont utilisé une approche faisant
l'énumération exhaustive de tous les déterminants potentiels ou encore se sont
appuyées sur l'un des deux modèles du stress occupationnel les plus influents,
nommément le modèle demandes-contrôle (Karasek, 1979) et le modèle de
déséquilibre efforts-récompenses (Siegrist, 1996), mettant ainsi l'accent sur un
nombre restreint de facteurs de stress bien spécifiques qui ne sont pas nécessairement
pertinents pour toutes les occupations, niveaux de responsabilités et contextes de
travail (Bakker & Demerouti, 2007).
L'introduction du modèle demandes-ressources (DR) a permis de progresser
sur le plan théorique. En assumant que les facteurs de stress associés à l'épuisement
professionnel peuvent être classés dans deux catégories générales, soit les demandes
et les ressources, le modèle DR constitue un modèle dominant, à plus grande portée,
qui simplifie la réalité complexe des organisations d'une telle manière qu'il peut être
utilisé pour analyser différents contextes occupationnels et niveaux de responsabilités
dans l'emploi (Bakker & Demerouti, 2007). Bien que le modèle DR remette en
question les modèles influents du stress occupationnel en raison de sa simplicité
conceptuelle et de sa flexibilité, il ne tient compte que des effets directs des demandes
et des ressources sur les symptômes de l'épuisement professionnel sans reconnaître le
rôle médiateur des facteurs et processus mentaux à l'intérieur de l'individu. Le modèle
Chapitre 7: Discussion générale
154
DR ressemble de près au modèle stimulus-réponse du comportement humain sur
lequel se fonde la théorie béhavioriste. Le béhaviorisme est basé sur la proposition
que toutes les choses que les organismes font - incluant agir, penser et ressentir peuvent et doivent être considérées comme des comportements (Skinner, 1984).
Cependant,
les
béhavioristes
s'intéressent
uniquement
aux
comportements
observables qui, argumentent-ils, constituent la réalité empirique. Sans pour autant
nier la réalité des individus et des processus internes, les béhavioristes n'examinent
pas le rôle des construits psychologiques jugés inobservables et hypothétiques. Ainsi,
l'individu est considéré comme une boîte noire.
Avec le premier article, nous avions pour objectif d'enrichir le modèle DR et
son approche conceptuelle des environnements de travail en intégrant, dans ce cadre
théorique, des éléments de la personne, induits par l'environnement de travail,
susceptibles d'agir à titre de facteurs de stress internes et de jouer un rôle médiateur
dans les relations entre les facteurs de stress externes tels que perçus par les individus
et l'épuisement professionnel. Ainsi, nous avons proposé une nouvelle taxonomie de
facteurs de stress associés à la vie au travail qui établit une distinction entre, d'une
part, les demandes et les ressources externes situées dans l'environnement de travail
(i.e. liées au travail en tant que tel et au contexte d'exécution du travail) et localisées à
différents niveaux d'analyse et, d'autre part, les demandes et les ressources internes
situées à l'intérieur de l'individu mais induites par l'environnement de travail et
localisées à ces mêmes différents niveaux d'analyse. En élaborant une nouvelle
taxonomie de facteurs de stress, nous cherchions à clarifier comment les différents
facteurs de stress évoqués dans la littérature sont interreliés et contribuent au
Chapitre 7: Discussion générale
155
développement de l'épuisement professionnel. Les relations postulées entre les
demandes externes perçues et les demandes internes induites par l'environnement de
travail localisées au même niveau et entre les ressources externes perçues et les
ressources internes induites par l'environnement de travail localisées au même niveau
ont toutes été confirmées. De plus, les résultats tendent à supporter de manière
générale notre hypothèse à l'effet que les relations entre les facteurs de stress d'un
même type, i.e. soit des demandes ou des ressources, situés à un même niveau
d'analyse sont plus fortes que les relations entre les facteurs de stress d'un même type
situés à des niveaux d'analyse différents. Quant au présumé rôle médiateur joué par
les demandes et les ressources internes induites par l'environnement de travail dans
les relations entre les facteurs de stress externes perçus et l'épuisement professionnel,
six des huit tests effectués ont confirmé la présence d'un effet de médiation. Enfin, les
résultats suggèrent que les demandes localisées au niveau de la tâche ou du rôle et les
ressources localisées au niveau de l'organisation pourraient être les facteurs
contribuant le plus au développement de l'épuisement professionnel chez les
travailleurs du milieu hospitalier. Cependant, puisque nous avons validé notre
taxonomie avec seulement un exemple de facteur de stress associé à chaque catégorie,
des études additionnelles sont nécessaires afin de confirmer ces résultats.
Un deuxième postulat du modèle DR qui nous suggéra une autre question de
recherche pour cette thèse est la conception de l'épuisement professionnel en tant
qu'état dysfonctionnel persistant qui résulte d'une exposition prolongée au stress
chronique. Avec le deuxième article, nous avions pour objectif d'examiner dans
quelle mesure l'articulation d'une vision organisationnelle et la promotion de valeurs
Chapitre 7: Discussion générale
156
organisationnelles liées à l'excellence, telles que la qualité et l'innovation, incitent les
employés en processus d'épuisement professionnel à choisir le présentéisme plutôt
que l'absentéisme pour cause de santé au cours de la prochaine année. Nous avons
argumenté que la vision et les valeurs organisationnelles peuvent accroître la
motivation des employés en processus d'épuisement professionnel à se présenter au
travail et qu'il est possible de concevoir le présentéisme comme le signe d'un niveau
d'engagement élevé et comme un comportement de citoyenneté organisationnelle.
Toutefois, ces ressources situées à un niveau d'analyse plus macro constituent des
caractéristiques de l'environnement de travail qui risquent de prolonger d'autant plus
l'exposition des travailleurs au stress chronique et donc de contribuer au
développement de l'épuisement professionnel. En effet, les résultats de notre étude
révèlent que l'articulation de la vision organisationnelle et la promotion de la qualité
et de l'innovation, telles que perçues par les employés en processus d'épuisement
professionnel, les incitent à choisir le présentéisme plutôt que l'absentéisme pour
cause de santé au cours de la prochaine année. Conséquemment, le présentéisme est
susceptible d'aggraver les symptômes d'épuisement professionnel et d'impacter de
manière plus conséquente la performance au travail pour ensuite donner lieu à un
accroissement des épisodes d'absences à court terme au cours de l'année subséquente.
De fait, notre étude appuie fortement la proposition à l'effet qu'une diminution de la
performance globale au travail, qui dit-on signalerait le présentéisme, est un
précurseur de l'absentéisme pour cause de santé.
Finalement,
le
troisième
article
démontre
comment
l'épuisement
professionnel, en tant qu'état dysfonctionnel persistant, peut persuader le personnel
Chapitre 7: Discussion générale
157
infirmier de quitter sa profession. Le stress chronique se caractérise par une
diminution significative et constante des ressources. La mobilisation des énergies
personnelles et d'autres ressources pour faire face aux demandes chroniques induit un
phénomène de perte en spirale (“loss spiral” phenomenon) qui affecte les ressources
en réserve (Hobfoll & Shirom, 2000). Après un certain temps, le personnel infirmier
aura souffert d'une grande perte en termes de ressources, tout comme les
établissements de santé. Selon la théorie du capital humain (Becker, 1993), la santé
est une ressources productive au même titre que le savoir et les compétences. Ainsi,
dans l'économie actuelle, les organisations aux prises avec une épidémie d'épuisement
professionnel sont elles-mêmes engagées dans un processus d'épuisement
professionnel, perdant des ressources humaines critiques alors qu'elles ont à faire face
à des demandes toujours plus grandes. Face aux établissements de santé
dysfonctionnels, l'ensemble de la société est d'autant plus à risque de souffrir d'une
perte importante de ressources dans le contexte actuel de pénurie du personnel
infirmier. En effet, les résultats de notre étude suggèrent que l'épuisement
professionnel joue un rôle important dans la décision de quitter la profession
infirmière, de par son impact sur les troubles psychosomatiques et sur l'engagement
professionnel. Afin de retenir le personnel infirmier au sein de la profession, les
résultats indiquent que ses tâches et son rôle devraient être restructurés afin de réduire
la charge de travail et accroître le sens du travail. Ces deux facteurs contribuant à
l'épuisement professionnel sont certes reliés: la surcharge de travail empêche le
personnel infirmier d'accomplir son travail relationnel auprès des patients et ainsi
d'assumer son rôle de soignant (“care giver”) tel qu'il le conçoit, en accord avec les
principes de la médecine holistique.
Chapitre 7: Discussion générale
158
Pour clore cette discussion générale, nous souhaitons partager un certains
nombre de réflexions quant aux limites du modèle DR sur les plans théorique,
pratique et méthodologique. En premier lieu, le fait que le modèle DR n'approche
l'épuisement professionnel que depuis les perspectives individuelle, interpersonnelle
et organisationnelle peut être vu comme une limitation. La propagation de
l'épuisement professionnel dans le secteur des services de la santé et des services
sociaux suggère qu'il existe des facteurs structurels plus profondément ancrés aux
niveaux sociétal et culturel contribuant aussi à l'épuisement professionnel (Schaufeli
& Enzmann, 1998). Dans une prochaine étude, il serait pertinent d'examiner les
causes des facteurs de stress liés à la vie au travail et de considérer les variables plus
macro liées à l'environnement externe de l'organisation (i.e. économie, technologie,
main-d'œuvre, valeurs et attitudes, lois) qui ont un impact sur la gestion des
ressources humaines. Le fait de ne pas considérer les interrelations entre les facteurs
sociaux et culturels à portée plus large, un environnement organisationnel spécifique
et l'expérience subjective du travailleur peut expliquer l'échec des interventions mises
de l'avant par les organisations pour prévenir l'épuisement professionnel
(cf.
Ruotsalainen, Serra, Marine, & Verbeek, 2008).
En second lieu, nous avons déjà soulevé le fait que le modèle DR ne fait pas
de distinction entre les facteurs de stress externes et les facteurs de stress internes
induits par l'environnement de travail. Par conséquent, le modèle DR considère que
les facteurs de stress externes et internes interviennent au même stade du processus
d'épuisement professionnel. Autrement dit, il ne tient compte que des effets directs
des demandes et des ressources sur les symptômes de l'épuisement professionnel sans
Chapitre 7: Discussion générale
159
reconnaître le rôle médiateur des facteurs et processus mentaux à l'intérieur de
l'individu. Cette limitation a des conséquences sur le plan pratique puisque le modèle
DR dans sa forme originale n'indique pas aux dirigeants et gestionnaires vers quels
facteurs de stress ils devraient orienter leurs efforts afin de prévenir l'épuisement
professionnel. Or, tel qu'argumenté dans le premier article, les facteurs de stress
externes sont les premiers responsables de la relation problématique entre la personne
et son environnement (stress). Il serait donc inefficace de la part des organisations de
ne chercher qu'à influencer directement les états psychologiques de leurs employés
par des interventions centrées sur l'individu.
En troisième lieu, bien que le modèle DR attribue des rôles bien différenciés
aux demandes et aux ressources, nous avons pu observer, dans le premier article, une
très forte association entre deux facteurs de stress internes associés aux interactions
sociales, soit entre le sentiment d'être supporté par le supérieur et celui d'être traité
injustement par lui sur le plan interpersonnel. Or, il y a lieu de se demander si les
concepts de support et de justice interpersonnelle sont bien indépendants l'un de
l'autre. Autrement dit, l'employé qui ne se sent pas supporté par son supérieur se sentil automatiquement traité injustement par lui sur le plan interpersonnel (et vice
versa)? Il est certain que le manque de support d'un supérieur peut être dû à plusieurs
facteurs (ex: à sa surcharge de travail, à son style de gestion) et non pas uniquement à
un manque de considération envers l'employé. Néanmoins, si ce raisonnement fait du
sens, nous sommes forcée de constater, face aux relations observées, que l'employé
fait peu de distinction entre les deux concepts et que le manque de support est pour
lui nécessairement une forme d'injustice interpersonnelle. Par contre, nous avons pu
Chapitre 7: Discussion générale
160
observer, dans le troisième article, que le fait d'inclure dans notre modèle de
recherche des facteurs de stress fortement reliés sur le plan conceptuel (ex: sens du
travail et impact de la tâche) peut introduire un problème de multicolinéarité
produisant une instabilité des coefficients estimés. Pour des raisons de parcimonie,
nous serions portés à former un construit de second-ordre avec ces facteurs de stress,
en autant qu'ils jouent le même rôle. L'inconvénient de cette procédure est une perte
d'information quand aux facteurs de stress spécifiques principalement en cause dans
le développement de l'épuisement professionnel. D'ailleurs, une limite importante sur
le plan pratique de nombreuses études menées sur le modèle DR est d'avoir examiné
l'impact des demandes et des ressources traitées en tant que facteurs de second-ordre
sur les symptômes d'épuisement professionnel.
En quatrième lieu, un postulat important du modèle DR est que deux
processus distincts sont en cause dans le développement de l’épuisement
professionnel (cf. Schaufeli & Bakker, 2004). Le processus énergétique origine des
demandes qui ont des effets principaux linéaires sur la dimension épuisement du
syndrome. Le processus motivationnel origine des ressources qui ont des effets
principaux linéaires sur la dimension cynisme du syndrome. Ainsi, les prédictions du
modèle DR sont établies sur la base d'un modèle additif qui stipule que lorsque les
demandes et les ressources sont élevées, le niveau d'épuisement sera élevé et le degré
de cynisme sera faible. Or, les recherches menées par Warr (1987) suggèrent que
certaines caractéristiques de l'environnement de travail, notamment des ressources
telles que le contrôle et le support mais aussi des demandes comme la surcharge
quantitative et le conflit de rôle, engendreraient des impacts négatifs sur la santé
Chapitre 7: Discussion générale
161
lorsqu’elles sont présentes à l’excès. Dans une étude basée sur le modèle de Karasek
(1979), Van der Doef, Maes et Diekstra (2000) ont également observé des effets
curvilinéaires pour le contrôle, le support et les demandes.
Par ailleurs, face aux recherches démontrant la présence d'effets d'interaction
suivant une fonction multiplicative entre les demandes et les ressources sur les
symptômes d'épuisement professionnel, Bakker et Demerouti (2007) ont révisé
récemment les prédictions du modèle DR. Ainsi, lorsque les demandes et les
ressources sont élevées, le niveau d'épuisement sera moyen, et non élevé tel que le
prédit le modèle additif, et le degré de cynisme sera faible. Lorsque les demandes et
les ressources sont faibles, le niveau de cynisme sera moyen, et non élevé tel que le
prédit le modèle additif, et le degré d'épuisement sera faible. Quant aux prédictions
concernant les situations de déséquilibre entre les demandes et les ressources, elles
demeurent les mêmes que celles basées sur le modèle additif. Par ailleurs, Bakker et
Demerouti (2007) soulèvent la possibilité que les effets d'interaction observés entre
les demandes et les ressources suivent une fonction proportionnelle plutôt que
multiplicative. De fait, l'utilisation du ratio entre les demandes et les ressources pour
prédire les degrés d'épuisement et de cynisme s'avèrerait justifiée sur le plan
théorique puisque l'une des fonctions des ressources assumée par le modèle DR est de
réduire directement les demandes. Ainsi, les symptômes d'épuisement professionnel
s'intensifieraient au fur et à mesure que la proportion des demandes rencontrées par
les ressources diminuerait. Les nouvelles prédictions du modèle DR établies sur la
base des études mettant en évidence des effets d'interaction entre les demandes et les
ressources nous amènent à conclure que les processus énergétique et motivationnel
Chapitre 7: Discussion générale
162
impliqués dans le développement de l'épuisement professionnel ne sont pas
indépendants. Autrement dit, on ne peut estimer l'influence des demandes sur
l'épuisement sans tenir compte du niveau des ressources, ni estimer l'influence des
ressources sur le cynisme sans tenir compte du niveau des demandes.
Évidemment, la nature exacte des effets des facteurs de stress (additifs ou
multiplicatifs) sur l'épuisement professionnel et, dans le cas du modèle additif, la
forme de la relation entre les facteurs de stress et l'épuisement professionnel, ont des
implications importantes sur le plan pratique. Les organisations contemporaines,
notamment celles du secteur de la santé, sont confrontées à des demandes croissantes.
Si le modèle additif s'avérait le plus robuste et que les effets curvilinéaires de
certaines ressources et demandes étaient confirmés, les dirigeants et gestionnaires
auraient alors à porter une attention particulière aux demandes grandissantes, en
limitant certaines ambitions organisationnelles, tout en maintenant un niveau de
ressources suffisamment élevé de façon à faciliter l'atteinte des objectifs de travail et
stimuler la croissance personnelle. Par contre, si le modèle multiplicatif (ou encore
proportionnel) s'avérait le plus robuste, les dirigeants et gestionnaires n'auraient alors
qu'à accroître les ressources disponibles sans avoir à modifier le niveau des
demandes. Toutefois, pour ce dernier scénario, comme les organisations disposent de
ressources limitées pour faire face à des demandes croissantes, elles auraient à faire
preuve d'une grande inventivité pour accroître leurs ressources.
CHAPITRE 8: CONCLUSION GÉNÉRALE
La plupart des livres de gestion du stress sont enclins à lister toutes les choses
que les individus font pour se créer eux-mêmes du stress. Maslach et Leiter (1997)
remettent en cause cette prémisse et tiennent les organisations responsables de
l'épuisement professionnel. Si l'épuisement professionnel, tel qu'ils le soutiennent, a
atteint aujourd'hui des proportions épidémiques chez les travailleurs nord-américains,
il est de fait difficile de maintenir que les travailleurs sont faibles et entièrement
responsables de leur sentiment de fatigue et de leur attitude cynique.
Il semble que les organisations contemporaines ne tiennent pas suffisamment
compte des ressources qu'elles possèdent - en particulier leur capital humain - et
tendent à se fixer des objectifs trop ambitieux. En mettant l'accent principalement sur
l'analyse externe des opportunités et des menaces dans un environnement sans cesse
en mouvement, les stratégies qu'elles implantent ne parviennent pas à éviter leurs
faiblesses internes. Il est possible d’établir un parallèle entre l’ascension d’Icare - qui
devait lui procurer une sorte de bien-être grisant - et les organisations qui sont sans
cesse en quête de dépassement - des autres et d'elles-mêmes. Extasié par son habileté
à voler, Icare s'éleva dans le ciel, mais il s'approcha trop près du soleil qui fit fondre
la cire de ses ailes retenant les plumes ensemble. Icare continua à battre des ailes mais
réalisa rapidement qu'il n'avait plus de plumes et qu'il n'agitait que ses bras
dépouillés. Ainsi, Icare plongea dans l'eau et y trouva la mort. Tout comme le mythe
d'Icare qui évoque le désire de l'Homme d'aller toujours plus loin au risque de devoir
se retrouver face à face avec sa condition de simple être humain, les organisations qui
Chapitre 8: Conclusion générale
164
aspirent continuellement vers de plus hauts sommets, cherchant toujours de nouvelles
façons de faire plus avec moins, s'exposent elles-mêmes ainsi que leur personnel à un
risque élevé d'épuisement professionnel.
Références
Adams, A., Lugsden, E., Chase, J., Arber, S., & Bond, S. (2000). Skill-mix changes and work
intensification in nursing. Work Employment and Society, 14, 541-555.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review
and recommended two-step approach. Psychological Bulletin, 103, 411-423.
Aiken, L. H., Clarke, S. P., & Sloane, D. M. (2002). Hospital staffing, organization, and quality
of care: Cross-national findings. International Journal for Quality in Health Care, 14(1),
5-13.
Argentero, P., Dell'Olivo, B., & Ferretti, M. S. (2008). Staff burnout and patient satisfaction with
the quality of dialysis care. American Journal of Kidney Diseases, 51(1), 80-92.
Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A metaanalytic review. British Journal of Social Psychology, 40, 471-499.
Aronsson, G., & Gustafsson, K. (2005). Sickness presenteeism: Prevalence, attendance-pressure
factors, and an outline of a model for research. Journal of Occupational Environmental
Medicine, 47, 958-966.
Aronsson, G., Gustafsson, K., & Dallner, M. (2000). Sick but yet at work. An empirical study of
sickness presenteeism. Journal of Epidemiological Community Health, 54, 502-509.
Aryee, S., Chen, Z. X., Sun, L.-Y., & Debrah, Y. A. (2007). Antecedents and outcomes of
abusive supervision: Test of a trickle-down model. Journal of Applied Psychology, 92(1),
191-201.
Aryee, S., & Tan, K. (1992). Antecedents and outcomes of career commitment. Journal of
Vocational Behavior, 40, 288-305.
Aubert, N., & De Gaulejac, V. (1991). Le coût de l'excellence. Paris: Seuil
Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re-examining the components of
transformational and transactional leadership using the Multifactor Leadership
Questionnaire. Journal of Occupational and Organizational Psychology, 72, 441-462.
Bagozzi, R.P. (1994). Principles of marketing research. Oxford: Blackwell Publishers.
Bagozzi, R. P., & Yi, Y. (1988). On the evaluation of structural equation models. Journal of the
Academy of Marketing Science, 16(1), 74-94.
Bakker, A.B., & Demerouti, E. (2007). The job demands-resources model: State of the art.
Journal of Managerial Psychology, 22(3), 309-328.
Bakker, A.B., & Demerouti, E. (2008). Towards a model of work engagement. Career
Development International, 13(3), 209-223.
Bakker, A. B., Demerouti, E., De Boer, E., & Schaufeli, W. B. (2003). Job demands and job
resources as predictors of absence duration and frequency. Journal of Vocational
Behavior, 62, 341-356.
Bakker, A. B., Demerouti, E., & Euwema, M. C. (2005). Job resources buffer the impact of job
demands on burnout. Journal of Occupational Health Psychology, 10(2), 170-180.
Bakker, A. B., Demerouti, E., Taris, T. W., Schaufeli, W. B., & Schreurs, P. J. G. (2003). A
multigroup analysis of the job demands-resources model in four home care organizations.
International Journal of Stress Management, 10(1), 16-38.
Bakker, A. B., Demerouti, E., & Verbeke, W. (2004). Using the job demands-resources model to
predict burnout and performance. Human Resource Management, 43(1), 83-104.
166
Barker, J.R. (1993). Tightening the iron cage: Concertive control in self-managing teams.
Administrative Science Quarterly, 38, 408-513.
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51(6), 1173-1182.
Bass, B. (1990). From transactional to transformational leadership: Learning to share the vision.
Organizational Dynamics, 18(3), 19-31.
Beck, A. (1978). Depression Inventory. Philadelphia: Center for Cognitive Therapy.
Becker, G. S. (1993). Human capital: A theoretical and empirical analysis, with special
reference to education (3rd ed.). Chicago: University of Chicago Press.
Beehr, T. A. (1985). The role of social support in coping with organizational stress. In T. A.
Beehr, & R. S. Bhagat (Eds.), Human stress and cognition in organizations (pp.375-398).
New York: Wiley.
Bentein, K., Vandenberg, R., Vandenberghe, C., & Stinglhamber, F. (2005). The role of change
in the relationship between commitment and turnover: A latent growth modeling
approach. Journal of Applied Psychology, 90(3), 468-482.
Bentler, P. M. (1992). On the fit of models to covariances and methodology to the Bulletin.
Psychological Bulletin, 112, 400-404.
Bentler, P. M., & Chou, C.-P. (1987). Practical issues in structural modeling. Sociological
Methods & Research, 16(1), 78-117.
Bergström, G., Bodin, L., Hagberg, J., Aronsson, G., & Josephson, M. (2009). Sickness
presenteeism today, sickness absenteeism tomorrow? A prospective study on sickness
presenteeism and future sickness absenteeism. Journal of Occupational and
Environmental Medicine, 51(6), 629-638.
Blais, M. R., Brière, N. M., Lachance, L., Riddle, A. S., & Vallerand, R. J. (1993). L'inventaire
des motivations au travail de Blais. Revue québécoise de psychologie, 14(3), 185-215.
Bourbonnais, R., Comeau, M., Viens, C., Brisson, C., Laliberté, D., Malenfant, R. et al. (1999).
La vie professionnelle et la santé des infirmières depuis la transformation du réseau de la
santé. Santé mentale au Québec, XXIV(1), 136-153.
Boxx, W. R., Odom, R. Y., & Dunn, M. G. (1991). Organizational values and value congruency
and their impact on satisfaction, commitment, and cohesion: An empirical examination
within the public sector. Public Personnel Management, 20(1), 195-205.
Burke, M. J., Brief, A. P., & George, J. M. (1993). The role of negative affectivity in
understanding relations. Journal of Applied Psychology, 78(3), 402-412.
Burke, P. J., & Reitzes, D. C. (1991). An identity theory approach to commitment. Social
Psychology Quarterly, 54, 239-251.
Burke, R. J., & Richardsen, A. M. (2000). Psychological burnout in organizations. In R.
Golembiewski (Ed.), Handbook of organizational behavior (pp. 327-363). New York:
M.A. Dekker.
Büssing, A., & Glaser, J. (2000). Four-stage process model of the core factors of burnout: The
role of work stressors and work-related resources. Work & Stress, 14(4), 329-346.
Buunk, B. P., de Jonge, J., Ybema, J. F., & de Wolff, C. J. (1998). Psychosocial aspects of
occupational stress. In P. J. D. Drenth, H. Thierry, & C. J. de Wolff (Eds.), Handbook of
work and organizational psychology (2nd ed., Vol. 2, pp. 145-182). Hove, UK:
Psychology Press.
Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic
concepts, applications, and programming. Hillsdale: Lawrence Erlbaum Associates.
167
Canadian Institute for Health Information. (2006). Workforce trends of registered nurses in
Canada, 2005. Retrieved June 26, 2009, from the Canadian Institute for Health
Information Web site: http://secure.cihi.ca/.
Canadian Institute for Health Information. (2008). Regulated nurses: Trends, 2003 to 2007.
Retrieved October 15, 2009, from the Canadian Institute for Health Information Web site:
http://secure.cihi.ca/.
Caplan, R. D., Cobb, S., French, J. R. P., Van Harrison, R., & Pinneau, S. R. (1980). Job
demands and worker health. Ann Arbor: University of Michigan, Institute of Social
Research.
Carlson, D. S., Kacmar, K. M., & Williams, L. J. (2000). Construction and initial validation of a
multidimensional measure of work-family conflict. Journal of Vocational Behavior, 56,
249-276.
Carlson, D. S., & Perrewé, P. L. (1999). The role of social support in the stressor-strain
relationship: An examination of work-family conflict. Journal of Management, 25(4),
513-533.
Carroll, J.B. (1961). The nature of data, or how to choose a correlation coefficient.
Psychometrika, 26, 347-372.
Carroll, J. F. X., & White, W. L. (1982). Theory building. Integrating individual and
environmental factors within an ecological framework. In W. S. Paine (Ed.), Burnout:
Research, theory, and intervention perspective (pp. 41-60). Beverly Hills: Sage
Caverley, N., Cunningham, J. B., & Mac Gregor, J. N. (2007). Sickness presenteeism, sickness
absenteeism, and health following restructuring in a public service organization. Journal
of Management Studies, 44(2), 304-319.
Cherniss, C. (1980). Staff burnout: Job stress in the human services. Beverly Hills: Sage.
Cheung, J. (2004). The decision process of leaving nursing. Australian Health Review, 28(3),
340-348.
Cohen, J., Cohen, P., West, S.G., & Aiken, L.S. (2003). Applied multiple regression/correlation
analysis for the behavioral sciences (3rd ed.). Hillsdale: Lawrence Erlbaum.
Collerette, P., & Schneider, R. (1996). Le pilotage du changement : une approche stratégique et
pratique. Sainte-Foy: Les Presses de l’Université du Québec.
Collins, J.C., & Porras, J.I. (1991). Organizational vision and visionary organizations. California
Management Review, 34(1), 30-52.
Colquitt, J. (2001). On the dimensionality of organizational justice: A construct validation
measure. Journal of Applied Psychology, 86(3), 386-400.
Cook, A., Hiroz, J., & Mildon, B. (2006). Magnet hospitals: Characteristics of magnet hospitals
(fact sheet I of II). Hamilton: Nursing Health Services Research Unit, a collaborative
project of the University of Toronto, Faculty of Nursing, and McMaster University,
School of Nursing.
Cooper, C. L., Dewe, P. J., & O'Driscoll, M. P. (2001). A special form of strain. Job related
burnout. In C.L. Cooper, P.J. Dewe, & M.P. O’Driscoll (Eds.), Organizational stress: A
review and critique of theory, research, and applications (pp. 79-115). Thousand Oaks:
Sage.
Cordes, C. L., & Dougherty, T. W. (1993). A review and an integration of research on job
burnout. Academy of Management Review, 18(4), 621-656.
Cropanzano, R., Rupp, D. E., & Byrne, Z. S. (2003). The relationship of emotional exhaustion to
work attitudes, job performance, and organizational citizenship behaviors. Journal of
Applied Psychology, 88(1), 160.
168
Darr, W., & Johns, G. (2008). Work strain, health, and absenteeism: A meta-analysis. Journal of
Occupational Health Psychology, 13(4), 293-318.
Darras, E., Dierckx de Casterlé, B., Milisen, K., Braes, T., Denhaerynck, K., Dierickx, K., et al.
(2001-2003). La profession infirmière en crise? Une recherche sur le concept de soi
professionnel des infirmiers. Louvain, Belgique: Université catholique de Louvain, Unité
des Sciences Hospitalières et Médico-sociales.
De Jonge, J., Dormann, C., Janssen, P. M. P., Dollard, M. F., Landeweerd, J. A., & Nijhuis, F. J.
N. (2001). Testing the reciprocal relationships between job characteristics and
psychological well-being: A cross-lagged structural equation model. Journal of
Occupational and Organizational Psychology, 74, 29-46.
De Jonge, J., Janssen, P. P. M., & Van Breukelen, G. J. P. (1996). Testing the demand-controlsupport model among health-care professionals: A structural equation model. Work &
Stress, 10(3), 209-224.
Demerouti, E., Bakker, A. B., & Bulters, A. J. (2004). The loss spiral of work pressure, workhome interference and exhaustion: Reciprocal relations in a three-wave study. Journal of
Vocational Behavior, 64, 131-149.
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demandsresources model of burnout. Journal of Applied Psychology, 86(3), 499-512.
Demerouti, E., Le Blanc, P.M., Bakker, A.B., Schaufeli, W.B., & Hox, J. (2009). Present but
sick: A three-wave study on job demands, presenteeism and burnout. Career
Development International, 14(1), 50-68.
Dew, K., Keefe, V., & Small, K. (2005). "Choosing" to work when sick: Workplace
presenteeism. Social Science & Medicine, 60, 2273-2282.
Doty, D. H., & Glick, W. H. (1998). Common methods bias: Does common methods variance
really bias results? Organizational Research Methods, 1(4), 374-406.
Doucet, O., Simard, G., & Tremblay, M. (2006). L'effet médiateur du support et de la confiance
dans la relation entre le leadership et l'engagement. Reims: XVIIe Congrès de
l’Association Francophone de Gestion des Ressources Humaines.
Edwards, J.R. (1992). A cybernetic theory of stress, coping, and well-being in organizations.
Academy of Management Review, 17(2), 238-274.
Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived organization
support. Journal of Applied Psychology, 71, 500-507.
Eisenstat, R. A., & Felner, R. D. (1983). Organizational mediators of the quality of care: Job
stressors and motivators in human service settings. In B. A. Farber (Ed.), Stress and
burnout in the human service professions (pp. 142-152). Elmsford: Pergamon Press.
Eisenstat, R. A., & Felner, R. D. (1984). Toward a differentiated view of burnout: Personal and
organizational mediators of job satisfaction and stress. American Journal of Community
Psychology, 12(4), 411-430.
Elstad, J. I., & Vabo, M. (2008). Job stress, sickness absence and sickness absenteeism in Nordic
elderly care. Scandinavian Journal of Public Health, 36, 467-474.
Fimian, M. J., Fasteneau, P. S., & Thomas, J. A. (1988). Stress in nursing and intentions of
leaving the profession. Psychological Reports, 62, 499-506.
Finegan, J. E. (2000). The impact of person and organizational values on organizational
commitment. Journal of Occupational and Organizational Psychology, 73, 149-169.
Firth, H., & Britton, P. (1989). Burnout, absence and turnover amongst British nursing staff.
Journal of Occupational Psychology, 62, 55-59.
169
Fishbein, M. (1967). Attitude and the prediction of behavior. In M. Fishbein (Ed.), Readings in
attitude theory and measurement (pp. 477-492). New York: John Wiley & Sons.
Fishbein, M., & Ajzen, T. (1975). Belief, attitude, intention, and behaviour: An introduction to
theory and research. Reading: Addison-Wesley.
Freudenberger, R., & Richelson, G. (1980). Burnout: The high cost of high achievement. New
York: Anchor Press/Doubleday.
Frone, M. R., Yardley, J. K., & Markel, K. S. (1997). Developing and testing an integrative
model of the work-family interface. Journal of Vocational Behavior, 50, 145-167.
Fu, C. K., & Shaffer, M. A. (2001). The tug of work and family. Direct and indirect domainspecific determinants of work-family conflict. Personnel Review, 30(5), 502-522.
Gagné, M., Senécal, C., & Koestner, R. (1997). Proximal job characteristics, feelings of
empowerment, and intrinsic motivation: A multidimensional model. Journal of Applied
Social Psychology, 27(14), 1222-1240.
Garman, A. N., Corrigan, P. W., & Morris, S. (2002). Staff burnout and patient satisfaction:
Evidence of relationships at the care unit level. Journal of Occupational Health
Psychology, 7(3), 235-241.
Godard, J., & Delaney, J. T. (2000). Reflections on the "high performance" paradigm's
implications for industrial relations as a field. Industrial & Labor relations Review, 53(3),
482-502.
Green, D. E., Walkey, F. H., & Taylor, A. J. W. (1991). The three-factor structure of the Maslach
Burnout Inventory: A multicultural, multinational confirmatory study. Journal of Social
Behavior and Personality, 6(3), 453-472.
Greenglass, E. R., & Burke, R. J. (2002). Hospital restructuring and burnout. Journal of Health
and Human Services Administration, 25(1/2), 89.
Griffeth, R.W., Hom, P.W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates
of employee turnover: Update, moderator tests, and research implications for the next
millennium. Journal of Management, 26(3), 463-488.
Grinyer, A., & Singleton, V. (2000). Sickness absence as risk-taking behaviour: A study of
organisational and cultural factors in the public sector. Health, Risk & Society, 2(1), 7-21.
Gutek, B. A., Searle, S., & Klepa, L. (1991). Rational versus gender role explanations for workfamily conflict. Journal of Applied Psychology, 76(4), 560-568.
Hackett, R.D., Bycio, P., & Guion, R.M. (1989). Absenteeism among hospital nurses: An
idiographic-longitudinal analysis. Academy of Management Journal, 32, 424-453.
Hackman, J. R. (1977). Work design. In J. R. Hackman, & J. L. Suttle (Eds.), Improving life at
work: Behavioral science approaches to organizational change (pp. 96-162). Santa
Monica: Goodyear Publishing Company.
Hackman, J. R., & Oldham, G. R. (1980). Work redesign. Reading: Addison-Wesley.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis
with readings (5th ed.). Upper Saddle River: Prentice-Hall.
Hakanen, J.J., Schaufeli, W.B., & Ahola, K. (2008). The job demands-resources model: A threeyear cross-lagged study of burnout, depression, commitment, and work engagement.
Work & Stress, 22(3), 224-241.
Hakkaart, L., & Essink, M.-L. (2000). The Health and Labour Questionnaire Manual.
Rotterdam: Institute for Medical Technology Assessment, Erasmus University Rotterdam.
Halbesleben, J. R. B., & Bowler, M. (2005). Organizational citizenship behaviors and burnout. In
D. L. Turnipseed (Ed.), Handbook of organizational citizenship behavior (pp. 399-414).
New York: Nova Science Publishers.
170
Halbesleben, J. R. B., & Buckley, M. R. (2004). Burnout in organizational life. Journal of
Management, 30(6), 859-879.
Halbesleben, J. R. B., Wakefield, B. J., Wakefield, D. S., & Cooper, L. B. (2008). Nurse burnout
and patient safety outcomes. Western Journal of Nursing Research, 30(5), 560-577.
Hammer, T. H., & Landau, J. (1981). Methodological issues in the use of absence data. Journal
of Applied Psychology, 66(5), 574-581.
Harrison, D.A., Johns, G., & Martocchio, J.J. (2000). Changes in technology, teamwork, and
diversity: New directions for a new century of absenteeism research. In G.R. Ferris (Ed.),
Research in personnel and human resources management (Vol. 18, pp. 43-91). Stamford:
Jai Press.
Hendrix, W. H., & Spencer, B. A. (1989). Development and test of a multivariate model of
absenteeism. Psychological Reports, 64, 923-958.
Hobfoll, S. E. (2002). Social and psychological resources and adaptation. Review of General
Psychology, 6(4), 307-324.
Hobfoll, S. E., & Freedy, J. (1993). Conservation of resources: A general stress theory applied to
burnout. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout:
Recent developments in theory and research (pp. 115-129). Washington: Taylor &
Francis.
Hobfoll, S. E., & Shirom, A. (2000). Conservation of resources theory. Applications to stress and
management in the workplace. In R. Golembiewski (Ed.), Handbook of Organizational
Behavior (2nd ed., pp. 57-80). New York: Dekker.
Hochwarter, W. A., Perrewé, P. L., Ferris, G. R., & Guercio, R. (1999). Commitment as an
antidote to the tension and turnover consequences of organizational politics. Journal of
Vocational Behavior, 55, 277-297.
Hockey, G. R. J. (1997). Compensatory control in the regulation of human performance under
stress and high workload: A cognitive-energetical framework. Biological Psychology, 45,
73-93.
Hu, L. T., & Bentler, P. M. (1995). Evaluating model fit. In R. H. Hoyle (Ed.), Structural
equation modeling: Concepts, issues, and applications (pp. 76-99). Thousand Oaks: Sage.
International Council of Nurses & Florence Nightingale International Foundation. (2006). The
global shortage of registered nurses: An overview of issues and actions. Geneva:
International Council of Nurses.
Jaccard, J. (2001). Interaction effects in logistic regression. Thousand Oaks: Sage Publications.
Jaccard, J., Turriso, R., & Wan, C.K. (1990). Interaction effects in multiple regression. Newbury
Park: Sage Publications.
Jackson, S.E., Schwab, R.L., & Schuler, R.S. (1986). Toward an understanding of the burnout
phenomenon. Journal of Applied Psychology, 71 (4), 630-640.
Jackson, S.E., Turner, J.A., & Brief, A.P. (1987). Correlates of burnout among public service
lawyers. Journal of Occupational Behavior, 8(4), 339-349.
Janssen, P. P. M., Peeters, M. C. W., De Jonge, J., Houkes, I., & Tummers, G. E. R. (2004).
Specific relationships between job demands, job resources and psychological outcomes
and the mediating role of negative work-home interference. Journal of Vocational
Behavior, 65, 411-429.
Jex, S. M. (1998). Stress and job performance: Theory, research and implications for managerial
practice. Thousand Oaks: Sage Publications.
171
Johansson, G., & Lundberg, I. (2004). Adjustment latitude and attendance requirements as
determinants of sickness absence or attendance. Social Sciences & Medicine, 58(10),
1857-1868.
Johns, G. (1995). Absenteeism. In N. Nicholson (Ed.), The Blackwell encyclopedic dictionary of
organizational behavior (pp. 1-3). Oxford: Blackwell.
Johns, G. (2001). The psychology of lateness, absenteeism and turnover. In N. Anderson, D. S.
Ones, H. K. Sinangil, & C. Viswesvaran (Eds.), Industrial, work and organizational
psychology (Vol. 2, pp. 232-252). London: Sage Publications.
Johns, G. (2008). Absenteeism and presenteeism: Not at work or not working well. In C.L.
Cooper, & J. Barling (Eds.), The Sage handbook of organizational behavior (pp.160-177).
London: Sage.
Johns, G. (2009). Absenteeism or presenteeism? Attendance dynamics and employee well-being.
In S. Cartwright, & C.L. Cooper, The Oxford handbook of organizational well-being
(pp.7-30). Oxford: Oxford University Press.
Kahn, R. L., & Byosiere, P. (1992). Stress in organizations. In M. D. Dunette, J. M. R. Hough, &
H. C. Triandis (Eds.), Handbook of industrial and organizational psychology (Vol. 4, pp.
571-650). Palo Alto: Consulting Psychologists Press.
Kalliath, T. J., O'Driscoll, M. P., Gillespie, D. F., & Bluedorn, A. C. (2000). A test of the
Maslach Burnout Inventory in three samples of health care professionals. Work & Stress,
14(1), 35-50.
Karasek, R. A., Jr. (1979). Job demands, job decision latitude, and mental strain: Implications for
job redesign. Administrative Science Quarterly, 24(2), 285-308.
Kerr, M., Laschinger, H., Severin, C., Almost, J., Thomson, D., O'Brien-Pallas, L., et al. (2002).
Monitoring the health of nurses in Canada. Ottawa: Canadian Health Services Research
Foundation.
Kirkpatrick, S. A., & Locke, E. A. (1996). Direct and indirect effects of three core charismatic
leadership components on performance and attitudes. Journal of Applied Psychology,
81(1), 36-51.
Kivimäki, M., Head, J., Ferrie, J. E., Hemingway, H., Shipley, M. J., Vahtera, J., et al. (2005).
Working while ill as a risk factor for serious coronary events: The Whitehall II study.
American Journal of Public Health, 95, 98-102.
Kohlberg, L. (1969). Stage and sequence: The cognitive-developmental approach to socialization.
In D. A. Goslin (Ed.), Handbook of socialization theory and research (pp. 347-480).
Chicago: Rand McNally.
Kopelman, R. E., Greenhaus, J. H., & Connolly, T. F. (1983). A model of work, family, and
interrole conflict: A construct validation study. Organizational Behavior and Human
Performance, 32(2), 198-216.
Kouzes, J. M., & Posner, B. Z. (1987). The leadership challenge. San Francisco: Jossey-Bass.
Krausz, M., Koslowsky, M., Shalom, N., & Elyakim, N. (1995). Predictors of intentions to leave
the ward, the hospital, and the nursing profession: A longitudinal study. Journal of
Organizational Behavior, 16(3), 277-288.
Landsbergis, P. A. (1988). Occupational stress among health care workers: A test of the job
demands-control model. Journal of Organizational Behavior, 9(3), 217-239.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer.
Lazarus, R. S., & Launier, R. (1978). Stress-related transactions between person and
environment. In L. A. Pervin, & M. Lewis (Eds.), Perspectives in interactional
psychology (pp. 287-327). New York: Plenum Press.
172
Lee, K., Carswell, J. J., & Allen, N. J. (2000). A meta-analytic review of occupational
commitment: Relations with Person- and work-related variables. Journal of Applied
Psychology, 85(5), 799-811.
Lee, R. T., & Ashforth, B. E. (1991). Work-unit structure and processes and job-related stressors
as predictors of managerial burnout. Journal of Applied Social Psychology, 21(22), 18311847.
Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three
dimensions of job burnout. Journal of Applied Psychology, 81(2), 123-133.
Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout and
organizational commitment. Journal of Organizational Behavior, 9(4), 297-308.
Leiter, M. P., & Maslach, C. (2004). Areas of worklife: A structured approach to organizational
predictors of job burnout. In P. L. Perrewé, & D. C. Ganster (Eds.), Research in
occupational stress and well being (Vol. 3, pp. 91-134). Amsterdam: Elsevier.
Lewin, K. (1951). Field theory in social science. New York: Harper & Row.
Liden, R. C., & Arad, S. (1996). A power perspective of empowerment and work groups:
Implications for human resources management research. In G. R. Ferris (Ed.), Research in
personnel and human resources management (Vol. 14, pp. 205-251). Greenwich: JAI
Press.
Lief, H. I., & Fox, R. C. (1963). Training for "detached concern" in medical students. In H. I.
Lief, V. F. Lief, & N. R. Lief (Eds.), The psychological basis of medical practice (pp. 1235). New York: Harper & Row.
Locke, E. A., & Latham, G. P. (1990). A theory of goal setting and task performance. Englewood
Cliffs: Prentice-Hall.
London, M. (1983). Toward a theory of career motivation. Academy of Management Review,
8(4), 620-630.
Luthans, F., & Youssef, C.M. (2007). Emerging positive organizational behavior. Journal of
Management, 33(3), 321-349.
March, J. G., & Simon, H. A. (1958). Organizations. New York: Wiley.
Martocchio, J. J., & Harrison, D. A. (1993). To be there or not to be there?: Questions, theories,
and methods in absenteeism research. In G. R. Ferris (Ed.), Research in personnel and
human resources management (pp. 259-328). Greenwich: JAI Press.
Maslach, C. (1976). Burned-out. Human Behavior, 16-22.
Maslach, C. (1998). A multidimensional theory of burnout. In C. L. Cooper (Ed.), Theories of
organizational stress (pp. 68-85). Oxford: Oxford University Press.
Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of
Occupational Behavior (pre-1986), 2(2), 99-113.
Maslach, C., & Jackson, S. E. (1986). Maslach Burnout Inventory Manual (2nd ed.). Palo Alto:
CPP.
Maslach, C., & Jackson, S. E. (1996). Maslach Burnout Inventory-Human Services Survey
(MBI-HSS). In C. Maslach, S. E. Jackson, & M. P. Leiter (Eds.), MBI Manual (3rd ed.).
Mountain View: CPP.
Maslach, C., & Leiter, M. (1997). The truth about burnout: How organizations cause personal
stress and what to do about it. San Francisco: Jossey-Bass.
Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burnout. Annual Review of
Psychology, 52, 397-422.
173
Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the antecedents, correlates,
and consequences of organizational commitment. Psychological Bulletin, 108(2), 171194.
Mattke, S., Balakrishnan, A., Bergamo, G., & S.J. Newberry. (2007). A review of methods to
measure health-related productivity loss. The American Journal of Managed Care, 13(4),
211-218.
McClelland, G.H., & Judd, C.M. (1993). Statistical difficulties of detecting interactions and
moderator effects. Psychological Bulletin, 114(2), 376-390.
McGrath, J. E. (1976). Stress and behavior in organizations. In M. D. Dunnette (Ed.), Handbook
of industrial and organizational psychology (pp. 1351-1396). Chicago: Rand McNally.
Meyer, J.P., & Allen, N.J. (1991). A three-component conceptualization of organizational
commitment. Human Resource Management Review, 1(1), 61-89.
Meyer, J. P., Allen, N. J., & Smith, C. A. (1993). Commitment to organizations and occupations:
Extension and test of a three-component conceptualization. Journal of Applied
Psychology, 78(4), 538-551.
Meyer, J. P., & Herscovitch, L. (2001). Commitment in the workplace. Toward a general model.
Human Resource Management Review, 11, 299-326.
Meyer, J. P., & Smith, C. (2000). HRM practices and organizational commitment: Test of a
mediation model. Canadian Journal of Administrative Sciences, 17(4), 319-331.
Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance,
and normative commitment to the organization: A meta-analysis of antecedents,
correlates, and consequences. Journal of Vocational Behavior, 61, 20-52.
Morrison, E. W., & Phelps, C. C. (1999). Taking charge at work: Extra-role efforts to initiate
workplace change. Academy of Management Journal, 42, 403-419.
Morrow, P. C. (1993). The theory and measurement of work commitment. Greenwich: JAI Press.
Motowidlo, S. J., Packard, J. S., & Manning, M. R. (1986). Occupational stress: Its causes and
consequences for job performance. Journal of Applied Psychology, 71, 618-629.
Mowday, R. T., Porter, L. W., & Steers, R. M. (1982). Employee-organizational linkages: The
psychology of commitment, absenteeism and turnover. New York: Academic Press.
Netemeyer, R. G., Johnston, M. W., & Burton, S. (1990). Analysis of role conflict and ambiguity
in a structural equation framework. Journal of Applied Psychology, 75, 148-157.
O'Reilly, C.A., & Chatman, J.A. (1996). Culture as social control: Corporations, cults, and
commitment. Research in Organizational Behavior, 18, 157-200.
Oswald, S. L., Mossholder, K. W., & Harris, S. G. (1994). Vision salience and strategic
involvement: Implications for psychological attachment to organizational and job.
Strategic Management Journal, 15(6), 477-489.
Paré, G., & Tremblay, M. (2007). The influence of high-involvement work practices, procedural
justice, organizational commitment and citizenship behaviors on information technology
professionals' turnover intentions. Group & Organization Management, 32(3), 326-357.
Peters, T., & Waterman, R. (1982). In search of excellence. New York: Harper and Row.
Podsakoff, P. M., MacKenzie, S. B., Morrman, R. H., & Fetter, R. (1990). Transformational
leader behaviors and their effects on follower's trust in leader, satisfaction, and
organizational citizenship behaviors. Leadership Quarterly, 1(2), 107-142.
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and
prospects. Journal of Management, 12(4), 531-544.
174
Preckel, D., Meinel, M., Kudielka, B.M., Haug, H.-J., & Fischer, J.E. (2007). Effort-rewardimbalance, overcommitment, an self-reported health: Is it the interaction that matters?
Journal of Occupational and Organizational Psychology, 80, 91-107.
Rhodes, S. R., & Doering, M. (1983). An integrated model of career change. Academy of
Management Review, 8(4), 631-639.
Rhodes, S. R., & Doering, M. M. (1993). Intention to change careers: Determinants and process.
The Career Development Quarterly, 42(1), 76-92.
Rizzo, J., House, R. J., & Lirtzman, S. I. (1970). Role conflict and role ambiguity in complex
organizations. Administrative Science Quarterly, 15, 150-163.
Rokeach, M. (1973). The nature of human values. New York: The Free Press.
Roussel, P., Durrieu, F., Campoy, É., & El Akremi, A. (2002). Méthodes d'équations
structurelles: Recherche et applications en gestion. Paris: Economica.
Ruotsalainen, J., Serra, C., Marine, A., & Verbeek, J. (2008). Systematic review of interventions
for reducing occupational stress in health care workers. Scandinavian Journal of Work,
Environment, and Health, 34(3), 169-178.
Ryan, R. M., & Deci, E. L. (2004). Overview of self-determination theory: An organismic
dialectical perspective. In E. L. Deci, & R. M. Ryan (Eds.), Handbook of selfdetermination research (pp. 3-27). Rochester: University of Rochester Press.
Salancik, G. R., & Pfeffer, J. (1978). A social information processing approach to job attitudes
and task design. Administrative Science Quarterly, 23, 224-253.
Schaufeli, W., & Bakker, A. B. (2004). Job demands, job resources, and their relationship with
burnout and engagement: a multi-sample study. Journal of Organizational Behavior,
25(3), 293-315.
Schaufeli, W., & Enzmann, D. (1998). The burnout companion to study and practice: A critical
analysis. London: Taylor & Francis.
Schaufeli, W., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996). Maslach Burnout InventoryGeneral Survey (MBI-GS). In C. Maslach, S. E. Jackson, & M. P. Leiter (Eds.), Maslach
Burnout Inventory Manual (3rd ed., pp. 19-26). Mountain View: CPP.
Schaufeli, W.B., & Taris, T.W. (2005). The conceptualization and measurement of burnout:
Common ground and worlds apart. Work & Stress, 19(3), 256-262.
Schaufeli, W., & Van Dierendonck, D. (1993). The construct validity of two burnout measures.
Journal of Organizational Behavior, 14(7), 631-647.
Schaufeli, W., Van Dierendonck, D., & Van Gorp, K. (1996). Burnout and reciprocity: Towards
a dual-level social exchange model. Work & Stress, 3, 225-237.
Schulz, R., Greenley, J. R., & Brown, R. (1995). Organization, management, and client effects on
staff burnout. Journal of Health and Social Behavior, 36, 333-345.
Schultz, A. B., & Edington, D. W. (2007). Employee health and presenteeism: A systematic
review. Journal of Occupational Rehabilitation, 17, 547-579.
Schonfeld, I. S. (1996). Relation of negative affectivity to self-reports of job stressors and
psychological outcomes. Journal of Occupational Health Psychology, 1(4), 397-412.
Shanafelt, T. D., Bradley, K. A., Wipf, J. E., & Back, A. L. (2002). Burnout and self-reported
patient care in an internal medicine residency program. Annals of Internal Medicine,
136(5), 358-367.
Shannon, H. S., Woodward, C. A., Cunningham, C. E., McIntosh, J., Lendrum, B., Brown, J. et
al. (2001). Changes in general health and musculoskeletal outcomes in the workforce of a
hospital undergoing rapid change: A longitudinal study. Journal of Occupational Health
Psychology, 6(1), 3-14.
175
Shirom, A. (2003). Job-related burnout: A review. In J. C. Quick, & L. E. Tetrick (Eds.),
Handbook of occupational health psychology (pp. 245-264). Washington: American
Psychological Association.
Shirom, A., Nirel, N., & Vinokur, A. D. (2006). Overload, autonomy, and burnout as predictors
of physicians' quality of care. Journal of Occupational Health Psychology, 11(4), 328342.
Shore, L.M., & Barksdale, K. (1998). Examining degree of balance and level of obligation in the
employment relationship: A social exchange approach. Journal of Organizational
Behavior, 19, 731-744.
Siegrist, J. (1996). Adverse health effects of high-effort/low reward conditions. Journal of
Occupational Health Psychology, 1(1), 27-41.
Simard, G., Chênevert, D., & Tremblay, M. (2000). Les antécédents de l’engagement
organisationnel : le cas d’une organisation publique du secteur de la santé. Paris: Xe
Congrès de l’Association Francophone de Gestion des Ressources Humaines, Paris.
Simoens, S., Villeneuve, M., & Hurst, J. (2005). Tackling nurse shortages in OCED countries.
Retrieved June 26, 2009, from the Organisation for Economic Co-operation and
Development Web site: http://www.oecd.org/.
Simpson, R. (1998). Presenteeism, power and organizational change: Long hours as a career
barrier and the impact on the working lives of women managers. British Journal of
Management, 9, S37-S50.
Skinner, B.F. (1984). The operational analysis of psychological terms. Behavioral and brain
sciences, 7(4), 547–581.
Snow, R. E., & Farr, M.J. (1987). Cognitive-conative-affective processes in aptitude, learning,
and instruction: An introduction. In. R.E. Snow, & M.J. Farr (Eds.), Aptitude, learning,
and instruction (Vol.3, pp.1-10). Hillsdale: Lawrence Erlbaum Associates.
Spector, P. E. (1986). Perceived control by employees: A meta-analysis of studies concerning
autonomy and participation at work. Human Relations, 39(11), 1005-1016.
Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions,
measurement, and validation. Academy of Management Journal, 38(5), 1442-1465.
Spreitzer, G. M. (1996). Social structural characteristics of psychological empowerment.
Academy of Management Journal, 39(2), 483-504.
Taris, T. W. (2006). Is there a relationship between burnout and objective performance? A
critical review of 16 studies. Work & Stress, 20(4), 316-334.
Taris, T.W., Le Blanc, P.M., Schaufeli, W.B., & Schreurs, P.J.G. (2005). Are there causal
relationships between the dimensions of the Maslach Burnout Inventory? A review and
two longitudinal tests. Work & Stress, 19(3), 238-255.
Tepper, B. J. (2000). Consequences of abusive supervision. Academy of Management Journal,
43(2), 178-190.
Testa, M.R. (1999). Satisfaction with organizational vision, job satisfaction and service efforts:
An empirical investigation. Leadership & Organization Development Journal, 20(3), 154161.
Theorell, T. (2003). To be able to exert control over one's own situation: A necessary condition
for coping with stressors. In J. C. Quick, & L. E. Tetrick (Eds.), Handbook of
occupational health psychology (pp. 201-219). Washington: American Psychological
Association.
176
Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An
"interpretive" model of intrinsic task motivation. Academy of Management Review, 15(4),
666-681.
Tsui, A.S., Pearce, J.L., Porter, L.W., & Tripoli, A.M. (1997). Alternatives approaches to the
employee-organization relationship: Does investment in employees pay off? Academy of
Management Journal, 40(5), 1089-1121.
Vandenberg, R. J., Richardson, H. A., & Eastman, L. J. (1999). The impact of high involvement
work processes on organizational effectiveness. Group & Organization Management,
24(3), 300-339.
Van der Doef, M., Maes, S., & Diekstra, R. (2000). An examination of the job demand-controlsupport model with various occupational strain indicators. Anxiety, Stress, & Coping, 13,
165-185.
Van Dyne, L., Graham, J. W., & Dienesch, R. M. (1994). Organizational citizenship behavior:
Construct redefinition, measurement, and validation. Academy of Management Journal,
37, 765-802.
Van Horn, J. E., Schaufeli, W., & Enzmann, D. (1999). Teacher burnout and lack of reciprocity.
Journal of Applied Social Psychology, 29, 91-108.
Voydanoff, P. (2004). The effects of work demands and resources on work-to-family conflict and
facilitation. Journal of Marriage and Family, 66, 398-412.
Wallace, J.E., & Brinkerhoff, M.B. (1991). The measurement of burnout revisited. Journal of
Social Service Research, 14(1/2), 85-111.
Wandelt, M. A., Pierce, P. M., & Widdowson, R. R. (1981). Why nurses leave nursing and what
can be done about it. American Journal of Nursing, 72-77.
Wanous, J. P., Reichers, A. E., & Austin, J. T. (2000). Cynicism about organizational change.
Group & Organization Management, 25(2), 132-153.
Warr, P. B. (1987). Work, unemployment, and mental health. Oxford: Clarendon Press.
Warshaw, P.R., & Davis, F.D. (1985). Disentangling behavioral intentions and behavioral
expectations. Journal of Experimental Social Psychology, 21, 213-228.
Watson, D., Pennebaker, J. W., & Folger, R. (1987). Beyond negative affectivity: Measuring
stress and satisfaction in the workplace. Journal of Organizational Behavior
Management, 8(2), 141-158.
White, J. (1993). Changing labour process and the nursing crisis in canadian hospitals. Studies in
Political Economy, 40, 103-134.
White, J. (1997). Health care, hospitals, and reengineering: The Nightingales sing the blues. In A.
Duffy, D. Glenday, & N. Pupo (Eds.), Good jobs, bad jobs, no jobs, the transformation of
work in the 21st century (pp. 117-142). Toronto: Harcourt Brace.
Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as
predictors of organizational citizenship and in-role behaviors. Journal of Management,
17(3), 601-617.
Williams, L. J., & Anderson, S. E. (1994). An alternative approach to method effects by using
latent-variable models: Applications in organizational behavior research. Journal of
Applied Psychology, 79(3), 323-331.
Williams, L. J., Gavin, M. B., & Williams, M. L. (1996). Measurement and non-measurement
processes with negative affectivity and employee attitudes. Journal of Applied
Psychology, 81(1), 88-102.
177
Woodward, C. A., Shannon, H. S., Cunningham, C. E., McIntosh, J., Lendrum, B., Rosenbloom,
D. et al. (1999). The impact of re-engineering and other cost reduction strategies on the
staff of a large teaching hospital: A longitudinal study. Medical Care, 37(6), 556-569.
Xanthopoulou, D., Bakker, A.B., Demerouti, E., & Schaufeli, W.B. (2007). The role of personal
resources in the job demands-resources model. International Journal of Stress
Management, 14(2), 121-141.