personality and work stress: the role of five-factor model
TRANSCRIPT
Personality and Work Stress:
The Role of Five-Factor Model Traits and
Cynicism in Perceptions of Work Characteristics
Maria Törnroos
Unit of Personality, Work, and Health Psychology
Institute of Behavioural Sciences
University of Helsinki, Finland
Academic Dissertation to be publicly discussed,
by due permission of the Faculty of Behavioural Sciences
at the University of Helsinki, Auditorium XII, Unioninkatu 34,
on the 24th April 2015, at 12 o’clock
University of Helsinki
Institute of Behavioural Sciences
Studies in Psychology, 109, 2015
2
Supervisors:
Professor Mirka Hintsanen
Faculty of Education, University of Oulu, Oulu, Finland
Institute of Behavioural Sciences, University of Helsinki, Helsinki, Finland
Docent Taina Hintsa
Institute of Behavioural Sciences, University of Helsinki, Helsinki, Finland
Reviewers:
Professor Marianna Virtanen
Unit of Expertise in Work and Organizations, Finnish Institute of Occupational Health,
Helsinki, Finland
Professor Saija Mauno
School of Social Sciences and Humanities, University of Tampere, Tampere, Finland
Opponent:
Professor Hugo Westerlund
Stress Research Institute, Stockholm University, Stockholm, Sweden
ISSN-L 1798-842X
ISSN 1798-842X
ISBN 978-951-51-0903-3 (pbk.)
ISBN 978-951-51-0904-0 (PDF)
http://ethesis.helsinki.fi
Helsinki University Print
Helsinki 2015
3
CONTENTS
Abstract ........................................................................................................................ 5
Tiivistelmä ..................................................................................................................... 6
Acknowledgements ....................................................................................................... 7
List of original publications ............................................................................................ 8
Abbreviations ................................................................................................................ 9
1 Introduction .......................................................................................................... 10
1.1 Stress – a physiological and psychological concept .......................................... 10
1.2 Work stress – theories and methodological issues ............................................ 11
1.2.1 The demand-control model ......................................................................... 12
1.2.2 The effort-reward imbalance model ............................................................. 13
1.3 Personality ........................................................................................................ 15
1.3.1 The Five-Factor Model of personality .......................................................... 16
1.3.2 Cynicism ..................................................................................................... 16
1.4 Individual differences in perceptions of work stress ........................................... 17
1.4.1 Five-Factor Model traits and stressful work characteristics ......................... 18
1.4.2 Cynicism and job strain ............................................................................... 19
1.5 Gaps in previous research ................................................................................. 20
2 Aims of the study ................................................................................................. 22
3 Methods ............................................................................................................... 24
3.1 Participants ....................................................................................................... 24
3.1.1 Design of the Young Finns study ................................................................ 24
3.1.2 Sample selection of the present study ......................................................... 24
3.2 Measures .......................................................................................................... 26
3.2.1 Five-Factor Model personality traits and cynicism (Studies I, II and III) ....... 27
3.2.2 Effort-reward imbalance (Studies I, III and IV) ............................................. 27
3.2.3 Job strain (Studies II and III) ....................................................................... 28
3.2.4 Covariates................................................................................................... 29
3.3 Statistical analyses ............................................................................................ 30
4 Results ................................................................................................................ 33
4.1 Five-Factor Model personality traits and effort-reward imbalance (Study I) ........ 33
4.2 Five-Factor Model personality traits and job strain (Study II) ............................. 36
4.3 Cynicism and job strain (Study III) ..................................................................... 39
4
4.4 Longitudinal measurement invariance of the effort-reward imbalance scales
(Study IV) ................................................................................................................ 40
5 Discussion ........................................................................................................... 43
5.1 Main findings ..................................................................................................... 43
5.1.1 Five-Factor Model personality traits and work characteristics...................... 43
5.1.2 Bi-directional associations of cynicism and job strain .................................. 46
5.1.3 Longitudinal measurement invariance of the effort-reward imbalance scales
............................................................................................................................ 47
5.2 Methodological considerations .......................................................................... 48
5.3 Theoretical implications ..................................................................................... 52
5.4 Conclusions and practical implications .............................................................. 54
6 References .......................................................................................................... 57
She believed she could so she did.
5
Abstract The role of individual differences in perceptions of stress has long been recognized.
Despite this, the models that are used to measure stress at the workplace—the job strain
model and the effort-reward imbalance model—were developed to assess strenuous
work characteristics and their health effects, regardless of the individual. Because work
characteristics are usually measured using self-reports the measures cannot be
completely objective. The present study examined the susceptibility of the job strain
model and the effort-reward imbalance model to Five-Factor personality traits and
cynicism. In addition, this study tested the longitudinal measurement invariance of the
effort-reward imbalance scales. This study was part of the ongoing prospective,
population-based Young Finns study. The measurements for the present study were
carried out in 2001, 2007, and 2012. Five-Factor personality traits were assessed with a
questionnaire on the Five-Factor model, and cynicism was assessed with a scale derived
from the Minnesota Multiphasic Personality Inventory. Work characteristics were
measured with questionnaires on the job strain model and the effort-reward imbalance
model.
The results showed that high neuroticism was associated with higher job strain and
higher effort-reward imbalance and that high agreeableness was associated with lower
job strain and lower effort-reward imbalance. High extraversion, high openness, and
high conscientiousness were associated with lower job strain. Furthermore, high
conscientiousness was related to lower effort-reward imbalance only in men. High job
strain prospectively predicted higher cynicism six years later. The effort-reward
imbalance scales achieved strict longitudinal measurement invariance and showed
adequate criterion validity.
Although developed to measure the structural work environment, the job strain model
and the effort-reward imbalance model seem to be susceptible to Five-Factor
personality traits—especially to neuroticism and agreeableness. In addition, high job
strain seems to have far reaching consequences on cynical attitudes. Furthermore, the
results show that scores on effort-reward imbalance from different time points can
reliably be compared with each other. This study shows that organizations and
occupational health services should apply a more person-oriented approach to
increasing wellbeing at work.
6
Tiivistelmä Yksilöllisten taipumusten merkitys stressin kokemisessa on jo pitkään tunnistettu ja
esimerkiksi persoonallisuuden on todettu voivan vaikuttaa jokaiseen stressiprosessin
vaiheeseen, altistumisesta palautumiseen. Kuitenkin työn stressaavia tekijöitä mittaavia
malleja (työn vaatimukset–hallinta -malli sekä ponnistelu–palkitsevuus -malli)
kehitettiin kuormittavien työolojen arviointiin sekä työolojen terveysvaikutusten
selvittämiseksi, riippumatta yksilöstä. Tässä tutkimuksessa tarkasteltiin
persoonallisuuden ja kyynisyyden yhteyttä kokemukseen työn stressaavista tekijöistä.
Tämän lisäksi tutkittiin ponnistelu–palkitsevuus -mittarin pysyvyyttä yli ajan. Tutkimus
on osa Lasten ja nuorten sepelvaltimotautiriski (LASERI) tutkimusta, joka alkoi vuonna
1980 ja jonka seurantaa on toteutettu säännöllisin väliajoin. Tämän tutkimuksen
mittaukset olivat vuosilta 2001, 2007 sekä 2012. Persoonallisuutta, kyynisyyttä sekä
työn stressaavia tekijöitä mitattiin kyselylomakkeilla.
Tulokset osoittivat, että neuroottisuus oli yhteydessä kokemukseen korkeasta
työstressistä kun taas sovinnollisuus oli yhteydessä kokemukseen matalasta
työstressistä. Korkea ulospäinsuuntautuneisuus, korkea avoimuus sekä korkea
tunnollisuus olivat yhteydessä matalaan koettuun työstressiin (vain työn vaatimukset–
hallinta -malliin). Korkea tunnollisuus oli yhteydessä myös matalaan ponnistelu–
palkitsevuus -mallin mukaiseen työstressiin mutta vain miehillä. Tämän lisäksi työn
vaativuus–hallinta -mallin mukainen työstressi ennusti korkeampaa kyynisyyttä.
Tulokset osoittivat myös, että ponnistelu–palkitsevuus -mittari oli pysyvä yli ajan.
Työn stressaavia tekijöitä mittaavat mallit kehitettiin kuvaamaan kuormittavia
työoloja mutta tutkimuksemme tulokset näyttävät, että persoonallisuus vaikuttaa siihen,
kuinka stressaaviksi yksilö kokee työolot. Etenkin neuroottisuus sekä sovinnollisuus
ovat piirteitä, jotka vaikuttavat olevan tärkeitä työympäristön kokemiselle, riippumatta
siitä kumpaa mallia käytetään arvioimaan työoloja. Tämän lisäksi työoloilla vaikuttaa
olevan kauaskantoisia vaikutuksia kyynisiin asenteisiin. Tulokset tuovat myös tärkeätä
tietoa siitä, että ponnistelu–palkitsevuus -mittaria voidaan luotettavasti käyttää
tutkimuksiin työolojen muutoksesta. Tästä tutkimuksesta saatavaa tietoa voidaan
hyödyntää tutkimuksissa, joissa työn vaatimukset–hallinta -mallia sekä ponnistelu–
palkitsevuus -mallia käytetään arvioimaan työoloja. Lisäksi tätä tietoa voidaan
hyödyntää työpaikan stressi-interventioiden suunnittelussa ja kehittämisessä.
7
Acknowledgements First and foremost, I would like to thank my supervisors, Professor Mirka Hintsanen
and Docent Taina Hintsa, for providing excellent supervision and guidance. I am also
deeply appreciative for having had the opportunity to work in the research group of
Professor Emerita Liisa Keltikangas-Järvinen, for all the wisdom and advice she has
given me throughout this process, and for the possibility to utilize the Young Finns data.
I want to express my gratitude to Professor Saija Mauno and Professor Marianna
Virtanen for reviewing my thesis. Their invaluable comments and expertise
substantially helped me to improve my thesis. I would also like to thank the whole
Young Finns team, especially Professor Olli T. Raitakari and Professor Jorma Viikari,
for administering the data. All my co-authors also deserve a thanks for improving my
manuscripts and sharing their knowledge. I also want to thank the National Doctoral
Programme of Psychology for funding my thesis.
I am truly thankful for my friends and co-workers in the Personality and Well-being
research group. For me social support is as important as other work characteristics—if
not more important. A very special thanks goes to my dearest friends at work, Associate
Professor Markus Jokela, Doctor Christian Hakulinen and Jaakko Airaksinen. Words
cannot begin to describe what you mean to me and how grateful I am for your
friendship. Other special thanks, and an apology for being so neurotic, go to my room-
mates Doctor Kim Josefsson and Doctor Tom Rosenström—thank you for putting up
with me. I also want to thank Professor Marko Elovainio, Docent Laura Pulkki-Råback,
Doctor Sari Mullola and Doctor Päivi Merjonen for listening to my—sometimes trivial,
sometimes real—problems and supporting me on this journey. Additional thanks go to
Doctor Ilmari Määttänen, Ulla Pulkkinen, Satu Kumpulainen and Doctor Paula Virtala
for inspiring conversations and for your fun company.
Last, but definitely not least, I want to thank my friends and family. Mira, the best of
best, through rough and tough, and my parents Rose-Marie Ölander and Aimo Saarinen,
who provided me with everything I needed, and always kept the door open for me to
pursue whatever I want. Ironically, the thing I wanted wasn’t that far from home after
all, though the field of study is different than that of my mother’s. However, I couldn’t
have succeeded in anything without the love and support from my husband Jerker and
our two beautiful children. You are my biggest fans and the light of my life.
8
List of original publications
This thesis is based on the following publications:
I Törnroos, M., Hintsanen, M., Hintsa, T., Jokela, M., Pulkki-Råback, L., Kivimäki, M.,
Hutri-Kähönen, N., & Keltikangas-Järvinen, L. (2012) Personality traits of the five-
factor model are associated with effort-reward imbalance at work: a population-based
study. Journal of Occupational and Environmental Medicine, 54(7), 875‒880.
DOI:10.1097/JOM.0b013e31824fe0e4
II Törnroos, M., Hintsanen, M., Hintsa, T., Jokela, M., Pulkki-Råback, L., Hutri-
Kähönen, N., & Keltikangas-Järvinen, L. (2013) Associations between five-factor
model traits and perceived job strain: a population-based study. Journal of
Occupational Health Psychology, 18(4), 492–500. DOI:10.1037/a0033987
III Törnroos, M., Elovainio, M., Keltikangas-Järvinen, L., Hintsa, T., Pulkki-Råback,
L., Hakulinen, C., Merjonen, P., Theorell, T., Kivimäki, M., Raitakari, O. T., &
Hintsanen, M. (In Press) Is there a two-way relationship between cynicism and job
strain? Evidence from a prospective population-based study. Journal of Occupational
and Environmental Medicine.
IV Törnroos, M., Keltikangas-Järvinen, L., Hintsa, T., Hakulinen, C., Pulkki-Råback,
L., Jokela, M., Hutri-Kähönen, N., & Hintsanen, M. (2014) Longitudinal measurement
invariance of the effort-reward imbalance scales in the Young Finns study.
Occupational and Environmental Medicine, 71(4), 289–294. DOI:10.1136/oemed-
2013-101947
The articles are reprinted with the kind permission of the copyright holders.
9
Abbreviations
NEO-FFI Neuroticism, Extraversion, Openness, Five-Factor Inventory
NEO-PI Neuroticism, Extraversion, Openness, Personality Inventory
CFI Comparative Fit Index
RMSEA Root Mean Square Error of Approximation
BIC Bayesian Information Criterion
ERI Effort-reward imbalance
CI Confidence Interval
10
1 Introduction
1.1 Stress – a physiological and psychological concept
In everyday life, the word stress can mean a multitude of things (Väänänen, Anttila,
Turtiainen, & Varje, 2012), for example the stressors that elicit the stress response, the
physiological reaction to a stressor or the stress process. The original definition of stress
is that its function is to mobilize the body for fight or flight (Cannon, 1915). In modern
Western societies, this reaction is abundant, as most of the stressors we encounter do
not pose a physical threat. Instead, the very reaction that should protect us from physical
harm exposes us to physiological reactions in our body that, if prolonged, affects our
health in a negative way (Chandola, Heraclides, & Kumari, 2010).
In the literature, stress has been defined both as a physiological and as a
psychological process. Physiological stress refers to the result of any demand on the
body and the stress response is considered a necessary adaptive means for survival
(Selye, 1956). In contrast, psychological stress has been defined as a mismatch between
the demands of the environment and the individual’s resources, so that some individuals
are more vulnerable to stress than others (Lazarus & Folkman, 1984). Lazarus (1966)
has suggested that stress should not be treated as a variable, but as a rubric for many
variables and processes that consist of but are not restricted to stressors, stress reactions,
and outcomes.
If prolonged, both physiological and psychological stress may have adverse health
consequences—e.g. cardiovascular disease, mortality, and depression—and it has been
suggested that psychological stress affects health via physiological pathways (for
example by activating the autonomic nervous system), via health behaviour or via
different aspects of psychological well-being (e.g. anxiety, depression, and distress)
(Bonde, 2008; Chandola et al., 2010; Elovainio et al., 2013; Hjemdahl, Rosengren, &
Steptoe, 2011; Kivimäki et al., 2013; Steptoe & Kivimäki, 2012).
11
1.2 Work stress – theories and methodological issues
It was not until the late 20th century that work stress became an important occupational
health issue in the Western societies and was understood as a threat to employees’
health and productivity (Väänänen et al., 2012). The term work stress has been used in
the literature to depict strenuous working conditions and their consequences on health
and wellbeing. The European Union defines work-related stress as an employee’s
experience of not being able to cope with (or control) the demands from the work
environment (EU-OSHA, 2009). Work stress, therefore, does not constitute the stress
process—from exposure to recovery—but the stressors and reactions to them. Across
Europe, the most common causes of work-related stress are job insecurity and workload
(EU-OSHA, 2013). In 2005, 22% of the work-force in the 15 EU countries reported
experiencing work-related stress (EU-OSHA, 2009). In Finland, 28% of the work-force
reported having a mentally strenuous job in 2012 and 8% reported having stress
symptoms (Kauppinen et al., 2013). Consequently, a majority of employees consider
their work non-strenuous and for some, work promotes well-being (van der Noordt,
IJzelenberg, Droomers, & Proper, 2014). However, those who do experience their
working environment as being highly strenuous have an increased risk for detrimental
health outcomes such as coronary heart disease or depression (Bonde, 2008; Kivimäki
et al., 2012; Stansfeld, Shipley, Head, & Fuhrer, 2012; Steptoe & Kivimäki, 2012).
In order to conceptualize work stress and identify psychosocial risk factors at work,
several models have been developed. As the measurement of stress at work is not
confined to the biological markers of physiological stress (e.g. cortisol levels) but relies
mostly on assessing psychological processes linked to the social environment, use of the
term stress is often discouraged. Instead, researchers use the terms psychosocial risks at
work or stressful work characteristics. The current study focuses on the two most
prominent theoretical models on stressful work characteristics, the demand-control
model (Karasek, 1979) and the effort-reward imbalance model (Siegrist, 1996). Both
models are valid, reliable and they have been used extensively in occupational research
to investigate the association between work characteristics and health outcomes (de
Lange, Taris, Kompier, Houtman, & Bongers, 2003; Kivimäki et al., 2012; Stansfeld &
Candy, 2006; van Vegchel, de Jonge, Bosma, & Schaufeli, 2005).
12
1.2.1 The demand-control model
Karasek’s (1979) demand-control model, or the job strain model, is one of the most
frequently used theories on stressful work characteristics. The model defines work
stress as two-dimensional: a high level of psychological demands combined with a low
level of decision latitude constitutes the highest risk for job strain and stress-related
diseases (Belkic, Landsbergis, Schnall, & Baker, 2004; Haeusser, Mojzisch, Niesel, &
Schulz-Hardt, 2010; Kivimäki et al., 2012). The demand component refers to workload,
while control (decision latitude) refers to the control over pace, use of skills, and
decision authority an employee has (Karasek & Theorell, 1990). According to the
model, the demands act as a stressor and control can act as a buffer to alleviate the strain
caused by the demands (Karasek, 1979). Because stress is not measured directly (but
through demand and control) the use of the word job strain instead of work stress, is
encouraged when using the job strain model to depict stressful work characteristics
(Karasek, 1979).
In order to explore the different effects of the interaction between demands and
control on health, their interaction coefficient—job strain—has in the work stress
literature been calculated using several different formulations, such as linear (job
demands - job control), quotient (job demands / job control), and multiplicative terms
(job demands x job control) (Landsbergis, Schnall, Warren, Pickering, & Schwartz,
1994; Schnall & Landsbergis, 1994). If the linear and quotient term show significant
associations, the effect of demand and control on job strain are additive (i.e. job strain
decreases by increasing control or decreasing demands) but if the multiplicative
interaction term is significant, its effect is stronger than the two components’ alone. In
addition, the multiplicative interaction implies that control acts as a buffer against high
demands (van der Doef & Maes, 1998). The use of continuous variables, such as linear
and quotient term, has been recommended (MacCallum, Zhang, Preacher, & Rucker,
2002) and when using a linear term of job strain the contributions of job demands and
job control are equally weighed (Landsbergis et al., 1994). It has also been stated that if
the main effects for demand and control are found, the implications for job redesign are
essentially the same for the different job strain formulations (Karasek, 1989). The job
strain model has further been extended with a component of social support, which
interacts with demand and control so that the highest job strain, or iso-strain, is caused
13
by conditions where demands are high but control and social support are low (Johnson
& Hall, 1988). The iso-strain model has gained support in the literature, albeit to a
lesser extent than the job strain model, indicating that the measurement of social support
is not as straightforward as the measurement of demand and control (Haeusser et al.,
2010).
1.2.2 The effort-reward imbalance model
A more recent model on stressful work characteristics is the effort-reward imbalance
model, in which an individual experiences stress at work if the reciprocity between
efforts spent and rewards received is not fulfilled (Siegrist, 1996). Effort is interpreted
as the demands and obligations the employee is faced with and reward as the money,
esteem, and career opportunities (or job security) the employee subsequently expects,
not only from the employer but also from society at large (Siegrist, 1996). According to
the extrinsic effort-reward imbalance hypothesis, the combination of high effort and low
reward—effort-reward imbalance—increases the risk of poor health independently of
the risks associated with the each of the components alone (Siegrist, 1996). The
extrinsic effort-reward imbalance hypothesis has been studied extensively and most
studies support the notion that the lack of reciprocity between effort and reward is
associated with employee health and well-being (Backe, Seidler, Latza, Rossnagel, &
Schumann, 2012; Godin, Kittel, Coppieters, & Siegrist, 2005; Niedhammer, Tek,
Starke, & Siegrist, 2004; van Vegchel et al., 2005). Effort-reward imbalance has also
been linked to health risk behaviours such as alcohol intake or being overweight
(Siegrist & Roedel, 2006).
In addition to extrinsic effort and reward, the effort-reward imbalance model also
includes intrinsic overcommittment, characterized by high need for approval and
excessive work-related commitment (Siegrist, 1996; Siegrist et al., 2004). Two
additional hypotheses were developed based on overcommittment: the intrinsic
overcommittment hypothesis, where high overcommittment is a psychosocial risk factor
even in the absence of effort-reward imbalance, and the interaction hypothesis, where
the highest risk on health occurs when all three components (effort, reward, and
overcommittment) interact (Siegrist et al., 2004). The intrinsic overcommittment
hypothesis has gained some support whereas the interaction hypothesis has been
14
scarcely studied and the results have been mixed (van Vegchel et al., 2005). According
to the effort-reward imbalance model, most detrimental to health is when efforts are
high and subsequent rewards are low over a long period of time (Siegrist et al., 2004).
This state of prolonged imbalance is likely to occur because of 1) lack of alternatives in
the labour market or limited mobility, 2) strategic reasons, such as hope of promotion,
or 3) high overcommittment (Siegrist et al., 2004).
Although overlapping to some extent (for example demand and effort correspond
quite closely to each other), the job strain model and the effort-reward imbalance model
reflect slightly different aspects of the psychosocial working environment (Siegrist et
al., 2004; Tsutsumi & Kawakami, 2004). Where the job strain model emphasizes task-
level control, the effort-reward imbalance model emphasizes the rewards the employee
receives (Siegrist et al., 2004). Control and reward also differ in how they reflect justice
at work. Control is thought to be related to procedural justice (Theorell, 2003), whereas
reward is thought to reflect distributive justice (Siegrist, 2001). Another characteristic
that separates the two models is the personal component overcommittment in the
otherwise situational effort-reward imbalance model (Siegrist et al., 2004). In addition,
the two models differ in their power of explaining stress in different occupational
settings (Calnan, Wadsworth, May, Smith, & Wainwright, 2004). Effort seems to
explain more of perceived stress in managers and professionals than in other
occupations, while perceived stress in sales and machine operatives is best explained by
lack of control (Calnan et al., 2004). Furthermore, the effort-reward imbalance model is
thought to be sensitive to changes in the labour market and to reflect the predominant
macro-economic labour-market situation whereas the job strain model focuses on work
place characteristics (Siegrist et al., 2004). Due to their similarities and differences, both
models seem to be suitable tools for measuring stressful work characteristics in the
current global economy characterized by increased insecurity in the labour market.
The surveillance of changes in work-related stress is important because these
changes might reflect reactions to organizational restructuring or outcomes of stress
interventions (Bourbonnais, Brisson, & Vezina, 2011; Limm et al., 2011). The analysis
of change in work characteristics relies on the invariance of the scales used and the
invariance of the scales allows researchers to make comparisons between constructs
over time, knowing that the operationalization of the construct has not changed (Schmitt
15
& Kuljanin, 2008; Vandenberg & Lance, 2000). Due to the changing world of work and
the sensitivity of the effort-reward imbalance model to these changes in the labour
market (Siegrist et al., 2004), assessing its longitudinal measurement invariance is
integral. Some previous studies on longitudinal measurement invariance of the effort-
reward imbalance scales exist. In a Dutch panel study on 383 (first wave) and 267
(second wave) healthcare workers (80–90% women) with a 1 to 2 years follow-up, the
factor loadings of the effort-reward imbalance scales were found not to be invariant
over time (de Jonge, van der Linden, Schaufeli, Peter, & Siegrist, 2008). In contrast, a
Finnish study on 758 white-collar professionals (14–17% women) showed that the
effort-reward imbalance scales were invariant across time (Rantanen, Feldt, Hyvönen,
Kinnunen, & Mäkikangas, 2013). The mixed findings may be a result of differences in
the samples and further research is needed in order to establish that the scales measure
work characteristics in the same way (i.e., the same latent variables) across repeated
measurement times.
1.3 Personality
Personality is a relatively stable, individual way of thinking, feeling and behaving
(Funder, 2012) and personality psychology has been proposed as the branch of
psychology concerned with identifying individual differences (Goldberg, 1981;
Norman, 1963). However, for a long time the field of personality research lacked tools
to identify personality—there was a need for a taxonomy that could be used to explore,
examine, and explain individual differences. The first taxonomy consisted of 16 primary
factors and 8 second-order factors, based on peer-ratings of college students (Cattell,
1948) and although this taxonomy was complex, it laid the ground for later trait
psychologists (Digman, 1990). Several researchers tried to replicate the factor structure
but they all came to the same conclusion: personality can be adequately described by
five superordinate dimensions (Fiske, 1949; Norman 1963; Tupes & Kristal, 1961).
Each of these dimensions—or traits—contains a large number of more specific
personality characteristics and captures the basic concepts of human personality.
16
1.3.1 The Five-Factor Model of personality
Continuing on the work of early trait psychologists Costa and McCrae (1985) developed
a questionnaire to assess the five dimensions that would explain individual differences.
The Five-Factor Model of personality has since become one of the most influential
taxonomies of personality traits in personality research. It specifies five personality
dimensions or traits: neuroticism, extraversion, openness, conscientiousness, and
agreeableness (McCrae & Costa, 1987). Neuroticism can be defined as a tendency to
worry, feelings of insecurity, and self-consciousness. Individuals high on the
neuroticism scale tend to be impulsive, experience more distress than others, and more
often resort to self-blame when confronted with negative feelings. Extraversion can be
defined as a tendency to be sociable, fun-loving, affectionate, friendly, and warm, and
extraverted individuals prefer the company of others to being alone. Openness is
characterized by originality, independence, and intellectual curiosity. Persons high on
the openness scale are full of ideas and values and may be seen by others as intelligent.
Conscientiousness can be described as a tendency to be well-organized, reliable, and
persevering, and persons high on the conscientiousness scale are more capable of self-
discipline than others and have an innate will to achieve. Lastly, agreeableness can be
characterized as a tendency to be sympathetic, and forgiving, and an agreeable person
usually trusts others and might therefore be taken advantage of more easily than persons
with low agreeableness (McCrae & Costa, 1987). The five dimensions have been shown
to be relatively enduring and stable over time, with normative personality change
occurring from young to old age (McCrae & Costa, 2003; Rantanen, Metsäpelto, Feldt,
Pulkkinen, & Kokko, 2007). However, there is some debate on what drives normative
personality change in adulthood—genes or social demands and experiences (McCrae &
Costa, 2003; Roberts & Mroczek, 2008; Specht, Egloff, & Schmukle, 2011).
1.3.2 Cynicism
Hostility is a personality construct that is closely related to many concepts of trait
psychology, such as neuroticism and agreeableness (Costa, Busch, Zonderman, &
Mccrae, 1986; Watson & Clark, 1992). While the Five-Factor Model traits are thought
of as major broad traits, hostility is described as a narrower facet or aspect of broader
17
traits. Cynicism represents the cognitive aspect of hostility and is defined as cynical and
mistrustful attitudes and the tendency to interpret other’s actions as offensive
(Greenglass & Julkunen, 1989; Smith, 1994). The development of cynical hostility may
be determined by both genes and environmental factors (Hakulinen et al., 2012;
Merjonen et al., 2011; Rebollo & Boomsma, 2006), including early childhood
experiences, low family socioeconomic status, parental Type A behaviour, and Type A
behaviour in childhood (Keltikangas-Järvinen & Heinonen, 2003). Cynicism is related
to many social problems, such as isolation (Vandervoort, 1999), depression (Nabi et al.,
2010), as well as somatic health problems, such as cardiovascular risk (Chida &
Steptoe, 2009). Therefore cynicism constitutes a public health risk and studying the
underlying factors and consequences is important.
1.4 Individual differences in perceptions of work stress
It has long been recognized that there are individual differences in stress reactivity and
stress responses (Lazarus & Folkman, 1984; Lazarus, 1999; Lovallo, 1997). According
to the classic theory by Lazarus and Folkman (1984), stress is caused by the interplay
between individual characteristics and stress factors, so that some people are more
vulnerable to stress than others. One of the most comprehensive individual
characteristics behind the stress experience is personality—an individual’s personality
may contribute to every stage of the stress process, i.e. exposure to the stressor,
appraisal of the stressor, coping, vulnerability to illness and disease, as well as response
to stress (Bolger & Zuckerman, 1995; Code & Langan-Fox, 2001).
In order to shed light on the relationship between personality and work
characteristics, several models have been proposed; the differential exposure model, the
differential reactivity model, the differential exposure-reactivity model, and the
outcome model (Bolger & Zuckerman, 1995; Kivimäki, 1996). According to the
differential exposure model, personality influences exposure to stressors (Bolger &
Zuckerman, 1995). Having a certain personality selects individuals to situations where
they are exposed to certain stressors, e.g. work characteristics. The differential reactivity
model, however, proposes that personality impacts on stress reactivity; a change in
stressors influences wellbeing differently for different personalities (Bolger &
18
Zuckerman, 1995). The differential exposure-reactivity model combines the two
previous models; personality impacts exposure and response to stressors (Bolger &
Zuckerman, 1995). In contrast to the other models, the outcome model suggests that
stressors influence personality, either directly or through wellbeing (Kivimäki, 1996).
Stress at work has often been examined through the conceptual framework of
environmental load, where occupational conditions at work are seen to cause stress. The
measures for assessing stressful working conditions—the job strain model and the
effort-reward imbalance model—were originally developed to depict structural aspects
of work (i.e. objective work characteristics). However, studies examining the effects of
the psychosocial working environment on employee health and wellbeing are often
based on self-reports. Self-reports, in turn, are vulnerable to individual dispositions,
such as personality, that may confound the associations of self-reported environmental
stress factors and experienced stress with related health outcomes. Thus, some people
may perceive, due to individual characteristics, that their environment is highly
stressful, as suggested by the differential exposure model. Population-based prospective
studies in Denmark (Ebstrup, Eplov, Pisinger, & Jorgensen, 2011) as well as in Finland
(Hintsa et al., 2010a; Hintsanen et al., 2011) have shown that personality and
temperament traits predict perceptions of job strain and effort-reward imbalance. In
addition, it has been shown that the direction of the association runs from personality to
stress, not the other way round (Sutin & Costa, 2010). However, personality has also
been found to moderate the association between work characteristics and health (Grant
& Langan-Fox, 2007; Moyle, 1995; Vahtera, Kivimäki, Uutela, & Pentti, 2000), which
might support the differential reactivity model. The role of personality in perceptions of
work characteristics defined by the job strain model and the effort-reward imbalance
model is, thus, still not clear.
1.4.1 Five-Factor Model traits and stressful work characteristics
Previous research has found associations between the Five-Factor personality traits and
different aspects of work. High neuroticism has been found to be associated with lower
work satisfaction, higher risk of burnout, and physical ill health, while the opposite has
been found for extraversion, conscientiousness, and agreeableness (Alarcon, Eschleman,
& Bowling, 2009; Grant & Langan-Fox, 2007; Judge, Heller, & Mount, 2002; Roberts,
19
Caspi, & Moffitt, 2003). In a recent study of middle-aged full-time employees
conducted in the United States, it was found that low neuroticism, high extraversion,
high conscientiousness, and high openness were related to greater decision latitude at
work (Sutin & Costa, 2010). However, the study did not find an association between
personality and psychological demands, and it did not assess job strain. A study by
Grant and Langan-Fox (2007) examined whether the Five-Factor personality traits
affect subjective strain at work and found that neuroticism, extraversion, and
conscientiousness were related to psychological strain. They did not, however, use the
job strain model as a measure of strain.
Despite previous evidence, however, little is known about the associations between
the Five-Factor Model traits and stressful work characteristics defined by the job strain
model and the effort-reward imbalance model. Some previous studies on conceptually
similar temperament traits imply that individual differences in perceptions of stressful
work characteristics do exist. For instance, studies on the association of temperament
with effort-reward imbalance and job strain have shown that negative emotionality, a
temperament trait conceptually close to neuroticism, predicts higher job strain and
effort-reward imbalance, whereas sociability, a temperament trait similar to
extraversion, predicts lower job strain and effort-reward imbalance (Hintsanen et al.,
2011). Studies on the trait negative affectivity, conceptually similar to neuroticism, have
also found direct associations with higher work stress (Moyle, 1995; Oliver, Mansell, &
Jose, 2010).
1.4.2 Cynicism and job strain
Studies suggest that cynicism is associated with work-related factors, such as
unemployment, unstable labour market prospects, and poor career achievement (Caspi,
Wright, Moffitt, & Silva, 1998; Hakulinen et al., 2013; Siegler et al., 2003). Although
studies on the association of cynicism with job strain are lacking, previous studies on
temperament and personality traits conceptually close to cynicism have been shown to
predict job strain (Hintsa et al., 2010a; Hintsa, Hintsanen, Jokela, Pulkki-Råback, &
Keltikangas-Järvinen, 2010b; Hintsanen et al., 2011). Based on these previous studies it
can therefore be assumed that cynicism would be associated with work characteristics
conceptualized by the job strain model.
20
According to the selection model, cynicism increases probability for higher exposure
to stress (Kivimäki et al., 2003; Smith, 1994). Because of the tendency of cynical
individuals to behave in an antagonistic and aggressive way, cynicism may produce
interpersonal conflict and lead to reduced social support, which, in turn, may increase
health risks (Smith, 1994). The vulnerability model suggests that cynical individuals are
more vulnerable to psychosocial risks than non-cynical (Kivimäki et al., 2003; Smith,
1994). Indeed studies have shown that cynical men experience less justice and social
support than non-cynical men (Elovainio, Kivimäki, Kortteinen, & Tuomikoski, 2001;
Elovainio, Kivimäki, Vahtera, Virtanen, & Keltikangas-Järvinen, 2003). In addition,
cynical individuals perceive their social environment more negatively than others
(Smith, 1994) and thus it is reasonable to assume that they make more negative and
extreme interpretations about their work environment.
The reverse direction of causality is also possible. Although cynicism is considered
to be a relatively stable trait in adulthood, mean level change over time is possible
(Hakulinen et al., 2014). According to the social context model adverse conditions, such
as psychosocial stress are antecedents of cynicism (Taylor, Repetti, & Seeman, 1997).
Previous studies have found that, in women, high workload predicts anger and
cynicism, defined as an employee’s hostile attitudes towards work situations
(Greenglass, Burke, & Moore, 2003; Greenglass, Burke, & Fiksenbaum, 2001). In
addition, psychosocial factors at work can cause burnout, which is often characterized
as cynicism, emotional exhaustion, and reduced professional efficacy (Lindblom,
Linton, Fedeli, & Bryngelsson, 2006; Schaufeli, Leiter, Maslach, & Jackson, 1996).
1.5 Gaps in previous research
The role of Five-Factor model traits and cynicism in perceptions of stressful work
characteristics is unclear. Although the structural work environment is important in
perceptions of work characteristics, individual dispositions, such as personality, also
play a part in what an individual is exposed to and how he or she perceives the working
environment. In order to target stress interventions and prevention appropriately,
identifying the role of individual traits, such as Five-Factor personality traits and
cynicism in perceptions of work characteristics is of great importance. In addition, the
21
research literature on the association between work characteristics and cynicism
suggests that the relationship might be bidirectional, but there are no prior studies that
would have investigated this.
Surveillance of change in work characteristics as a result of intervention or
organizational restructuring relies on the longitudinal measurement invariance of the
measures used. Previous research on the invariance of the effort-reward imbalance
scales is mixed, probably due to differences in the samples and an uneven distribution
of gender and occupational groups. Therefore, more research is needed using
population-based samples, a wider range of occupational groups, and an even gender
distribution. It is also important to examine whether effort-reward imbalance scales
have measurement invariance over longer time lags than four years. Examining the
susceptibility of the job strain model and the effort-reward imbalance model to
personality and establishing measurement invariance of the effort-reward imbalance
scales brings information that can be important for a large amount of organizational and
intervention studies.
22
2 Aims of the study
The first aim of the present study was to examine the associations of Five-Factor
personality traits with effort-reward imbalance and job strain. The Five-Factor Model of
personality defines five personality dimensions: neuroticism, extraversion, openness,
conscientiousness, and agreeableness. The second aim of this study was to prospectively
examine whether cynicism predicts job strain, or whether the association runs the other
way. The third aim of this study was to examine whether the effort-reward imbalance
scales are invariant over two measurement points five years apart. In addition, criterion
validity of the effort-reward imbalance scales was examined using a single-item
questionnaire on general stress. The specific research questions and hypotheses were as
follows:
1) Are the Five-Factor personality traits associated with job strain and effort-reward
imbalance?
Hypothesis 1a: high neuroticism is associated with higher effort-reward
imbalance, whereas high extraversion, high openness, high conscientiousness,
and high agreeableness are associated with lower effort-reward imbalance (Study
I).
Hypothesis 1b: high neuroticism is associated with higher job strain, whereas
high extraversion, high openness, high conscientiousness, and high agreeableness
are associated with lower job strain (Study II).
2) Does cynicism predict job strain, and is the association bi-directional?
Hypothesis 2: high cynicism predicts higher job strain and high job strain
predicts higher cynicism (Study III).
3) Are the effort-reward imbalance scales measurement invariant over two time-
points and do the effort-reward imbalance scales have criterion validity?
Hypothesis 3a: the effort-reward imbalance scales are invariant over time (Study
IV).
Hypothesis 3b: the effort-reward imbalance scales prospectively predict general
stress (Study IV).
23
The study variables at the different study phases are depicted in Table 1.
Table 1. Study variables at different phases of the study
Study phase
Study 2001 2007 2012
I and II
Five-Factor
personality traits
Job strain
ERI
III Cynicism
Job strain
Cynicism
Job strain
IV ERI
General stress
ERI
General stress
ERI = Effort-reward imbalance
24
3 Methods
3.1 Participants
3.1.1 Design of the Young Finns study
The sample for the present study was from the Cardiovascular Risk in Young Finns
study, or shortly the Young Finns study. The Young Finns study is an ongoing,
prospective population-based study, designed to study the risk factors of cardiovascular
diseases and their determinants in children and adolescents in Finland (Raitakari et al.,
2008; Åkerblom et al., 1991). The study was launched in 1978 and 1979 with two pilot
studies, and the first cross-sectional study was conducted in 1980. A total of 4320
participants from six age cohorts (3-, 6-, 9-, 12-, 15-, and 18-year olds) in the population
register of the Social Insurance Institution covering the entire geographic area of
Finland and nationally representative of various socioeconomic groups were initially
invited to the study in 1980. Of the invited, 3596 participants (83.2% response rate)
responded in the first study. The follow-ups have been conducted in 1983, 1986, 1989,
1992, 1997, 2001, 2007, and the latest follow-up in 2012.
3.1.2 Sample selection of the present study
The criteria for inclusion in studies I-IV were full-time work, no missing data in the
covariates and a maximum of 50% missing data in the study variables. The
measurements for study I and II were carried out in 2007, when 2058 participants
(57.2% response rate), aged 30 to 45 years, responded to the survey on the
psychological variables. In Study I, we included 1370 participants (Mage = 38 years)
with adequate information on Five-Factor personality traits and the components of the
effort-reward imbalance model. Attrition analyses showed that compared to the
excluded, the included participants in Study I were proportionally more often men than
women (72.5% vs. 62.5%, p < .001), and that the included participants were slightly
older than the excluded participants (37.99 vs. 36.78, p < .001). In comparison with the
excluded participants, the included participants had lower scores on neuroticism (2.33
vs. 2.53, p < .001) and higher scores on extraversion (3.46 vs. 3.38, p < .01) in addition
25
to experiencing higher effort (3.25 vs. 3.09, p < .001) and higher reward (3.76 vs. 3.65,
p < .01). Furthermore, the included participants had higher educational level (2.34 vs.
2.25, p = .001) and higher occupational status (2.18 vs. 1.93, p < .001) than the
excluded participants.
In Study II, we included 1372 participants (Mage = 38 years) with adequate
information on Five-Factor personality traits and the components of the job strain
model. The attrition analyses showed that as compared to the excluded participants the
included participants in Study II were proportionally more often men (44.8% vs. 33.3%,
p < .001) and the included participants were slightly older than the excluded participants
(37.99 vs. 36.80, p < .001). In comparison with the excluded participants, the included
participants had lower scores on neuroticism (2.33 vs. 2.54, p < .001) and higher scores
on extraversion (3.42 vs. 3.33, p = .001). Furthermore, the included participants had
higher educational level (2.33 vs. 2.25, p < .01) and higher occupational status (2.17 vs.
1.95, p < .001) than the excluded participants.
In Study III, measurements were carried out in 2001 (N = 2105, 58.5% response rate)
and 2007 when the participants responded to a survey on cynicism and components of
the job strain model. Based on inclusion criteria, we included 757 participants (399
women, 53%) in the structural equation models on the relationship between cynicism
and job strain (Mage in 2001 = 31.5 years). The attrition analyses showed that there were
proportionally more men in the included sample than in the excluded sample (47.3% vs.
41.7%, p = .009) and that the included participants were somewhat older (32.65 vs.
31.02, p < .001) than the excluded. Compared to the excluded, the included participants
had higher educational level (2.24 vs. 2.16, p = .001) and occupational status (1.99 vs.
1.87, p < .001) in 2001. In addition, the included participants reported having fewer
children in 2007 than the excluded (1.47 vs. 1.65, p = .005) and proportionally fewer of
the included had moved during the follow-up compared with the excluded participants
(58.6% vs. 65.4%, p = .001).
In Study IV, the data were collected in 2007 and 2012 (N = 1752, 48.7% response
rate). The number of participants varied according to the analyses; there were 1228 and
1177 participants in the analyses of invariance for effort and reward, respectively (Mage
= 38 years). In the regression analyses on the association of effort-reward imbalance
26
and its components with general stress, there were 1237 (unadjusted model) and 1083
(adjusted for occupational status and educational level) participants.
All participants gave written informed consent, and the study was approved by local
ethic committees.
3.2 Measures
The Cronbach’s alphas of the scales used in this study are shown in Table 2.
Table 2. Cronbach’s alphas for the scales in the study
2001 2007 2012
Studies I and II
Neuroticism
0.88
Extraversion
0.81
Openness
0.84
Conscientiousness
0.72
Agreeableness
0.80
Study III
Cynicism 0.79 0.83
Studies II and III
Demand 0.61 0.63
Control 0.86 0.87
Studies I and IV
Effort
0.76 0.76
Reward 0.82 0.83
27
3.2.1 Five-Factor Model personality traits and cynicism (Studies I, II and III)
In Studies I and II, personality was measured in 2007 using the Finnish version of the
NEO-FFI (Neuroticism, Extraversion, Openness, Five-Factor Inventory), which was
developed by Rantanen et al. (2007). The Finnish version used in this study contains 60
questions that are based on the questions from the original NEO-FFI (Costa & Mccrae,
1989) as well as on questions from the Finnish version of the NEO-PI (Personality
Inventory). The original NEO-PI version was developed by Costa and McCrae (1985)
and translated and standardized by Pulver, Allik, Pulkkinen, and Hämäläinen (1995).
Some of the questions in the Finnish PI version are modified in order to better
correspond to non-Indo-European languages.
Neuroticism was measured with 12 questions (e.g. “I sometimes feel completely
worthless”), extraversion with 12 questions (e.g. “I want to be surrounded by other
people”), conscientiousness with 12 questions (e.g. “I work hard in order to accomplish
my goals”), openness with 12 questions (e.g. “I am intellectually very curious”), and
agreeableness with 12 questions (e.g. “I would rather cooperate than compete with
others”). The participants answered the questions on a scale from 1 (does not apply) to 5
(applies well). For those who had less than 50% missing values, a mean score was
calculated for each personality trait.
In Study III, cynicism was assessed in 2001 and 2007 with a 7-item cynicism scale
derived from the Minnesota Multiphasic Personality Inventory (e.g. “It is safer to trust
nobody”) (Comrey, 1957; Comrey, 1958). Cynicism represents the cognitive aspect of
hostility (Smith, 1994) and has been identified as the central dimension of hostility
(Greenglass & Julkunen, 1989). The answers were given on a scale from 1 (totally
disagree) to 5 (totally agree). The longitudinal measurement invariance of the cynicism
scale has been shown in the data previously (Hakulinen et al., 2014).
3.2.2 Effort-reward imbalance (Studies I, III and IV)
In Studies I and IV, effort was assessed with the five-item original questionnaire (e.g. “I
have constant time pressure due to a heavy work load”) and reward with 11 items from
the original scale (Siegrist & Peter, 1996). The reward questionnaire consisted of five
items measuring esteem (e.g. “I receive the respect I deserve from my superiors”), four
28
items measuring job promotion (e.g. “My job promotion prospects are poor”), and two
items measuring job security (e.g. “My job security is poor”). The responses for both
effort and reward were given on a scale from 1 (does not apply) to 5 (does apply).
Mean scores for effort and reward were calculated for those participants who had a
maximum of 50% missing values. Effort-reward imbalance was calculated by dividing
the mean scores of the effort component by the mean scores of the reward component
(Siegrist et al., 2004). A higher value of the continuous effort-reward imbalance
variable indicated a higher imbalance (Siegrist et al., 2004). In Study I, a logarithmic
transformation was made to the effort-reward imbalance variable to correct for
skewness and kurtosis.
In Study IV, general stress in 2012 was used as an outcome criterion for effort-
reward imbalance and its components. General stress was measured in 2007 and 2012
by a one-item questionnaire on stress symptoms from the Occupational Stress
Questionnaire (“Stress means a situation in which a person feels tense, restless, nervous
or anxious or is unable to sleep at night because his/her mind is troubled all the time. Do
you feel this kind of stress these days?”) (Elo, Leppänen, Lindström, & Ropponen,
1992). Response was given on a scale from 1 (not at all) to 5 (very much). The validity
of the single-item measure has been shown previously (Elo, Leppänen, & Jahkola,
2003).
3.2.3 Job strain (Studies II and III)
In Studies II and III, job demands was measured in 2001 and 2007 using a three-item
scale from the Occupational Stress Questionnaire (e.g. “Does your work require you to
work fast?”), developed by the Finnish Institute of Occupational Health (Elo et al.,
1992). These three items correspond to the items in Karasek’s Job Content
Questionnaire (1985). The responses were given on a scale from 1 (never) to 5 (all the
time). Job control was in Studies II and III, measured in 2001 and 2007 using 9 items
from the Job Content Questionnaire (e.g. “In my work, I am allowed to make a lot of
decisions”) (Karasek, 1985). The response scale was from 1 (disagree) to 5 (agree).
Mean scores for demand and control were calculated only for those participants with a
maximum of 50% missing values.
29
Job strain in 2007 was in Study II calculated using three different formulations, linear
term (job demands - job control), quotient term (job demands / job control), and
multiplicative interaction term (job demands x job control). The multiplicative
interaction term was formed using centralized values for demands and control. When
analyzing the multiplicative term’s association, the main effects of job demands and job
control are also controlled for. In Study III, job strain in 2001 and 2007 were calculated
using the linear term.
3.2.4 Covariates
In addition to age and gender, the covariates in our study were educational level and
occupational status, since they are potential confounders (Brunner et al., 2004). Age was
entered as a covariate because recent studies have shown that age is associated with
both personality and work characteristics (Anusic, Lucas, & Donnellan, 2012; De Lange
et al., 2010; Shultz, Wang, Crimmins, & Fisher, 2010; Soto & John, 2012; Specht et al.,
2011). Educational level was classified as (1) low (comprehensive school), (2)
intermediate (secondary education), or (3) high (academic; graduated from a
polytechnic or a university). Occupational status was based on the Central Statistical
Office of Finland: (1) manual, (2) lower non manual, and (3) upper non manual. The
occupational status of entrepreneurs was determined based on their educational level
(low, intermediate, and high education corresponding to manual, lower non-manual, and
upper non-manual respectively) and was coded accordingly into occupational status
categories. Educational level and occupational status were treated as categorical
variables when calculating correlations but dummy-coded in the regression analyses.
To take into account major life events during the follow-up we also included the
following covariates in study III: marital status, which was coded (0) single, divorced or
widowed and (1) married or co-habiting; number of children (range 0-10); moving to a
new address during the follow-up, which was coded (0) no change in address and (1)
change in address during the follow-up. The participants’ address information was
derived from the Population Registry. In addition, both job strain and cynicism have in
several previous studies been shown to predict depressive symptoms later in life
(Bonde, 2008; Heponiemi et al., 2006; Nabi et al., 2010; Virtanen et al., 2015). To
30
examine this effect in Study III, we performed additional mediation analyses, with
depression in 2007 as an outcome. Depressive symptoms were measured using Beck’s
Depression Inventory II (BDI-II) (α = 0.92) (Beck, Steer, & Brown, 1996). The
participants answered the 21 statements on a scale from 1 (totally disagree) to 5 (totally
agree).
3.3 Statistical analyses
The associations between work characteristics and personality traits in Studies I and II
were examined by linear regression analyses. The analyses were performed separately
for each Five-Factor trait using two models. In Study I, the first model was adjusted for
age and the second model was additionally adjusted for education and occupational
status. In Study II, the first model was adjusted for age, gender, education, and
occupational status. In the second model all Five-Factor personality traits were entered
simultaneously into the analysis. The latter model enabled us to examine the association
of a trait with the outcome measures so that the other traits were held constant. The
analyses in Study I were performed separately for women and men as there were
significant gender interactions for neuroticism in relation to effort (p = .006) and reward
(p = .027), for extraversion in relation to reward (p = .001), and for conscientiousness in
association with reward (p < .001) and effort-reward imbalance (p = .017). In Study II,
gender interactions for neuroticism in relation to job control (p = .002) and job demands
(p = .008) were found, but as the results for women and men were very similar on these
variables (job control: β = -.240, p < .001 for women and β = -.385, p < .001 for men
when adjusted for age; job demands: β = .259, p < .001 for women and β = .119, p <
.001 for men when adjusted for age), the genders were combined.
In Study III a cross-lagged structural equation models on the relationship between the
latent construct cynicism and the observed variable job strain from both study phases
(2001 and 2007) was fitted to the data (Figure 1). The model was adjusted for age,
gender, educational level, change in occupational status, change in marital status,
having children during the follow up, and moving to another address during the follow-
up. The structural equation model allowed for correlation between measurement errors
over time. In addition, three items in the cynicism scale had high correlations (r > .50)
31
in addition to being conceptually similar to each other, and their errors were therefore
allowed to correlate within time-points. Additional mediation analysis in Study III, for
the association between job strain and depression with cynicism as a mediator (N =
750), was conducted using the sgmediation package in Stata, which utilizes Sobel-
Goodman’s test of mediation. To account for the effect of the covariates, we first
regressed all the covariates of the study on the job strain variable and the depression
variable. The mediation analysis was then performed using the residuals of job strain
and depression.
In Study IV, longitudinal measurement invariance was assessed with a series of
confirmatory factor analyses differentiating four types of measurement invariance:
configural invariance, weak (metric) invariance, strong (scalar) invariance, and strict
(residual) variance (see, e.g., Schmitt & Kuljanin, 2008). In order to test the invariance
of effort and reward, we tested the two dimensions separately following the theoretical
model which states that effort-reward imbalance is a division of effort and reward
(Siegrist et al., 2004). Reward was examined in two ways, as a first-order factor (i.e.,
sum score of all the components) and as a second-order factor consisting of the first-
order factors esteem, promotion, and security. Because the same individual items were
measured at both time points, their residual variances were allowed to correlate in all
models. In the effort scale, two sets of two items had a correlation over 0.4 in addition
to being theoretically very similar. Thus, we allowed them to correlate in all models. In
the first-order reward scale, 7 items were allowed to correlate as shown in Figure 2. The
structural model for reward as a second-order factor is shown in Figure 3. Factor
loadings of the effort and reward items in 2007 are shown in Figure 2 (effort and first-
order reward) and Figure 3 (second-order reward). The items are labelled according to
the original labelling (Siegrist et al., 2004). In addition, the associations of effort-reward
imbalance and its components in 2007 with general stress in 2012 were examined by
linear regression analysis adjusting for age, gender, occupational status, educational
level, and general stress at baseline.
In Studies III and IV, model fit was evaluated based on Comparative Fit Index (CFI)
and Root Mean Square Error of Approximation (RMSEA) index. RMSEA is not
affected by model complexity and CFI is independent of sample size (Cheung &
Rensvold, 2002). CFI values above .95 and RMSEA values below .08 indicate good fit.
32
Bayesian Information Criterion (BIC) was in Study IV used to compare models: the
lower the BIC, the better the model fit.
33
4 Results
Sample characteristics of the study are shown in Table 3.
4.1 Five-Factor Model personality traits and effort-reward
imbalance (Study I)
In Study I, women were slightly older (p = .049), had higher occupational status (p <
.001), and higher educational level (p = .05) than men. Women had higher scores on all
personality traits (p < .001) in addition to experiencing fewer possibilities of promotion
(p = .046) than men. The results for the linear regression analyses on the association
between personality traits and effort-reward imbalance are shown in Table 4. In both
genders, high neuroticism was associated with high effort-reward imbalance (β = .416, p
< .001) and high agreeableness with low effort-reward imbalance (β = -.197, p < .001)
when age, educational level, and occupational status were controlled for. Among
women, conscientiousness was not related to effort-reward imbalance, whereas high
conscientiousness was associated with low effort-reward imbalance in men (β = -.155, p
< .001). There were no associations between extraversion or openness and effort-reward
imbalance in either gender.
High neuroticism and high extraversion were associated with high effort while high
agreeableness was associated with low effort. In addition, high openness and high
conscientiousness were associated with high effort in women. As for the reward
component, high neuroticism was related to low reward and high extraversion,
conscientiousness, and agreeableness were related to high reward. No associations
between openness and reward were found in either gender.
34
Table 3. Descriptive statistics of the study variables
Study I Study II Study III Study IV
Women Men
Variable Mean / Count
SD / %
Mean / Count
SD / %
Mean / Count
SD / %
Mean / Count
SD / %
Mean / Count
SD / %
Demographics
Age -07 38.22 4.99 37.68 5.00 37.99 4.99 32.65 4.86 37.64 5.02
Educational level -07
low 25 3.3 28 4.6 53 3.9 29 3.8 88 4.6
intermediate 436 57.6 373 60.8 811 59.1 467 61.7 1127 59.1
high 296 39.1 212 34.6 508 37.0 261 34.5 691 36.3
Occupational status -07
manual 197 26.0 245 40.0 445 32.4 254 33.6 556 33.4
lower non manual 169 22.3 84 13.7 252 18.4 137 18.1 313 18.8
upper non manual 391 51.7 284 46.3 675 49.2 366 48.3 795 47.8
Five-Factor Model traits
Neuroticism 2.45 0.66 2.19 0.62 2.33 0.65
Extraversion 3.52 0.52 3.38 0.53 3.42 0.54
Openness 3.25 0.52 3.08 0.51 3.17 0.52
Conscientiousness 3.78 0.54 3.61 0.55 3.71 0.55
Agreeableness 3.75 0.48 3.56 0.48 3.69 0.49
Cynicism -01
2.69 0.68
Cynicism -07
2.50 0.72
Job strain and its components
Job demands -01
Job demands -07
2.93 0.65
Job control -01
Job control -07
3.77 0.69
Job strain (linear term) -01
-0.94 0.90
Job strain (linear term) -07
-0.84 0.85 -0.83 0.85
ERI and its components
Effort -07 3.28 0.83 3.21 0.78
3.21 0.81
Effort -12
3.26 0.81
Reward: esteem -07 3.79 0.73 3.75 0.75
Reward: promotion -07 3.49 0.73 3.57 0.76
Reward: security -07 3.97 0.96 3.98 0.86
Reward -07 3.75 0.61 3.77 0.62
3.70 0.61
Reward -12
3.71 0.64
ERI -07 -0.07a 0.14a -0.08a 0.13a
0.89 0.28
ERI -12
0.91 0.30
General stress -07
2.34 0.95
General stress -12 2.39 0.94
ERI = Effort-reward imbalance
a Logarithmically transformed
35
Table 4. Standardized linear regression coefficients for the association between Five-factor model traits and effort-
reward imbalance and its components
Women (n=757)
Effort Reward Effort-reward
imbalance
β ΔR² β ΔR² β ΔR²
Neuroticism Model 1 .19*** .04
-.44*** .19 .40*** .16
Model 2 .22*** .05
-.44*** .19
.42*** .17
Extraversion Model 1 .11** .01
.28*** .08
-.05 .00
Model 2 .08* .01
.27*** .07
-.07 .00
Openness Model 1 .13*** .02
.06 .00
.09* .01
Model 2 .09* .01
.05 .00
.06 .00
Conscientiousness Model 1 .10** .01
.20*** .04
-.02 .00
Model 2 .08* .01
.20*** .04
-.03 .00
Agreeableness Model 1 -.03 .00
.28*** .08
-.17*** .03
Model 2 -.07 .00
.28*** .07
-.20*** .04
Men (n=613)
Effort Reward Effort-reward
imbalance
β ΔR² β ΔR² β ΔR²
Neuroticism Model 1 .05 .00
-.51*** .26 .33*** .11
Model 2 .10* .01
-.49*** .23
.36*** .12
Extraversion Model 1 .22*** .05
.43*** .19
-.06 .00
Model 2 .17*** .03
.40*** .15
-.08 .01
Openness Model 1 .11** .01
.12** .02
.03 .00
Model 2 .05 .00
.05 .00
.02 .00
Conscientiousness Model 1 .08* .01
.41*** .17
-.15*** .02
Model 2 .07 .01
.40*** .17
-.16*** .02
Agreeableness Model 1 -.09* .01
.32*** .10
-.26*** .07
Model 2 -.14*** .02 .29*** .08 -.28*** .07
* Significant at the 0.05 level (two-tailed) Model 1 - adjusted for age
** Significant at the 0.01 level (two-tailed) Model 2 - adjusted for age, educational level, and
occupational status
*** Significant at the 0.001 level (two-tailed)
ΔR² is calculated for the personality trait
36
4.2 Five-Factor Model personality traits and job strain (Study II)
The results for linear regression analyses in Study II on the associations of personality
traits with job strain are shown in Table 5. When the traits were examined separately
and age, gender, educational level, and occupational status were controlled for, high
neuroticism was associated higher job strain (linear term; β = .397, p < .001).
Extraversion (β = -.263, p < .001), openness (β = -.090, p = .001), conscientiousness (β
= -.196, p < .001), and agreeableness (β = -.149, p < .001) were all inversely associated
with job strain. Furthermore, high neuroticism and high openness were associated with
higher demands while high agreeableness was associated with lower demands. High
neuroticism was related to lower control while high extraversion, high openness, high
conscientiousness, and high agreeableness were all related to higher control.
In general the associations remained fairly similar when the traits were examined
simultaneously, except for the associations of extraversion with demands (β = .149, p <
.001), which was not significant when extraversion was examined separately. In
addition the associations of extraversion with job strain, openness with demands, and
agreeableness with control, demand, and job strain attenuated to non-significance.
When the personality traits were examined separately, the associations with job strain
calculated as quotient term remained fairly similar to the results with the linear term. All
other associations remained significant, with the exception of openness. The
multiplicative term showed a significant association with extraversion. The other traits’
associations diminished to non-significance. Thus, extraversion is associated with
perceptions of job strain both when demand and control are non-independent—i.e. the
level of one depends on the level of the other—and when demand and control are
treated as independent factors. The other traits are associated with perceptions of job
strain only when demand and control are independent of each other—as in the linear
term.
Results for the analyses between simultaneously entered personality traits and
different formulations of job strain showed that quotient term job strain had a significant
association only with neuroticism and that multiplicative interaction term showed a
significant association with extraversion and agreeableness. Thus, when the effect of the
other traits is controlled for, extraversion and agreeableness are not associated with
37
perceptions of job strain if demand and control are treated as independent factors, but
are associated with job strain when demand and control are dependent of each other.
38
Table 5. Results of linear regression analyses on Five-Factor Model personality traits, job strain and its components
Demands
Control
Strain (Linear term)
Strain (Quotient term)
Straina (Multiplicative)
∆R2m
= .08
∆R2m
= .14
∆R2m
= .16
∆R2m
= .13
∆R2m = .00
β ΔR² β ΔR² β ΔR² β ΔR² β ΔR²
Neuroticism Model 1 .24*** .06
-.26*** .07
.40*** .15
.37*** .13
-.01 .00
Model 2 .31*** .06
-.14*** .01
.35*** .08
.34*** .07
-.01 .00
Extraversion Model 1 .01 .00
.33*** .11
-.26*** .07
-.23*** .05
.01** .00
Model 2 .15*** .01
.21*** .03
-.06 .00
-.05 .00
.01* .00
Openness Model 1 .06* .00
.17*** .03
-.09*** .01
-.05 .00
.00 .00
Model 2 .03 .00
.11*** .01
-.07** .00
-.04 .00
.00 .00
Conscientiousness Model 1 -.03 .00
.21*** .04
-.20*** .04
-.16*** .02
.00 .00
Model 2 .03 .00
.10*** .01
-.06* .00
-.03 .00
-.00 .00
Agreeableness Model 1 -.09*** .01
.10*** .01
-.15*** .02
-.13*** .02
-.01 .00
Model 2 -.03 .00 -.03 .00 .01 .00 .01 .00 -.01* .00
Values are standardized beta coefficients, their p-values and coefficients of determination.
Model 1: Individually entered traits adjusted for age, gender, occupational status, and educational level
Model 2: Simultaneously entered traits adjusted for age, gender, occupational status, educational level, and personality traits
ΔR² is calculated for the personality trait
∆R2m is calculated for all the traits in the Five-Factor Model combined
a Additionally adjusted for demands and control
* Significant at the 0.05 level (two-tailed)
** Significant at the 0.01 level (two-tailed)
*** Significant at the 0.001 level (two-tailed)
39
4.3 Cynicism and job strain (Study III)
Model fit showed that the structural equation model of the association of cynicism with
job strain in Study III fitted the data well [χ2 (df) = 463.18 (231), CFI = .949, RMSEA =
.036]. All factor loadings for the items in the cynicism scale were significant at the .001
level and ranged between 0.31 and 0.76. Figure 1 depicts the standardized coefficients
of the association between cynicism and job strain. The results revealed that high job
strain (β = .08, p = .006) was associated with higher cynicism six years later. The
association was independent of age, gender, educational level, change in occupational
status, change in marital status, having children, and moving during the follow-up.
The results for the additional Sobel-Goodman meditation analysis showed that
cynicism mediated the relationship between job strain in 2001 and depression in 2007 (p
< .001). The total effect mediated by cynicism was 21.5%.
Figure 1. Cross-lagged structural equation model of cynicism
and job strain with standardized coefficients. Measurement
errors (not shown) are allowed to correlate over time. Adjusted
for age, gender, educational level, change in occupational
status, change in marital status, having children, and change in
address.
* p < .05 ** p < .01 *** p <.001
40
4.4 Longitudinal measurement invariance of the effort-reward
imbalance scales (Study IV)
Table 6 shows a summary of the fit indices for the invariance models in Study IV. The
results show that both the effort scale and the reward scale are invariant over the two
time points used in our study. For effort, the strict model fit the data best (RMSEA =
.048, CFI = .977, BIC = 33418.078). Both RMSEA and BIC values were lower in the
strict model than in the strong (RMSEA = .051, BIC = 33448.412) or weak (RMSEA =
.054, BIC = 33474.448) model. For first-order reward, all models showed adequate fit
but examination of the BIC revealed that the strict model (RMSEA = .056, CFI = .916,
BIC = 66166.819) fit the data better than the strong (BIC = 66222.601) or the weak
(BIC = 66227.953) model. Likewise, when reward was treated as a second-order factor
the strict model had the best fit (RMSEA = .052, CFI = .924, BIC = 66038.978) when
comparing the BIC to the strong (BIC = 66101.982) or the weak (BIC = 66097.109)
model.
The results for the associations of effort-reward imbalance and its components in
2007 with general stress in 2012 showed that high effort (β = .269, p < .001) and high
effort-reward imbalance (β = .294, p < .001) were associated with high general stress
five years later while high reward (β = -.129, p < .001) was associated with low general
stress five years later. These associations were not attenuated after adjusting for
educational level and occupational status. The results indicate that the effort-reward
imbalance scales show criterion validity with the general stress measure. Additionally
adjusting for baseline general stress decreased the estimates so that only high effort
prospectively predicted higher general stress (β = .072, p = .019).
41
Figure 2. Correlations between conceptually similar items and factor loadings in the effort and reward
scales in 2007 (all factor loadings shown were significant at p < .001).
Figure 3. Factor loadings of items in the second-order reward scale with first-order factors esteem,
promotion and security in 2007 (all factor loadings shown were significant at p < .001).
42
Table 6. Summary of goodness-of-fit for the invariance models.
χ2/ df CFI RMSEA BIC
Effort
Baseline model 126.28/ 25 0.978 0.057 33492.47
Weak 129.60/ 28 0.978 0.054 33474.45
Strong 139.13/ 33 0.977 0.051 33448.41
Strict 144.36/ 38 0.977 0.048 33418.08
Reward - first order
Baseline model 901.36/ 187 0.925 0.057 66268.55
Weak 924.40/ 196 0.923 0.056 66227.95
Strong 996.824/ 207 0.917 0.057 66222.60
Strict 1018.82/ 218 0.916 0.056 66166.82
Reward - second order
Baseline model 828.73/ 194 0.933 0.053 66146.43
Weak 857.19/ 205 0.931 0.052 66097.11
Strong 939.84/ 216 0.924 0.053 66101.98
Strict 954.62/ 227 0.924 0.052 66038.98
CFI = Comparative Fit Index, RMSEA = Root Mean Square Error of
Approximation, BIC = Bayesian Information Criterion
43
5 Discussion
5.1 Main findings
5.1.1 Five-Factor Model personality traits and work characteristics
In the present study all personality traits of the Five-Factor Model explained differences
in perceived job strain while neuroticism, conscientiousness, and agreeableness
explained differences in effort-reward imbalance. High neuroticism was associated with
higher perceived job strain and effort-reward imbalance, whereas high agreeableness
was associated with lower perceived job strain and effort-reward imbalance. In addition,
high extraversion and high openness were associated with lower job strain. High
conscientiousness was associated with lower job strain in both genders and lower effort-
reward imbalance in men. Openness was not associated with effort-reward imbalance in
either gender. These associations were largely independent of age, educational level,
and occupational status. Thus, the hypotheses set for Studies I and II were supported
with the exception of openness. The results indicate that personality traits are associated
with the experience of work characteristics.
High neuroticism was in the present study associated with higher effort, lower
rewards, higher demands, lower control, and consequently higher effort-reward
imbalance and higher job strain. These results are supported by a recent study on Italian
police officers in which high neuroticism was associated with higher job strain and
higher effort-reward imbalance (Garbarino, Cuomo, Chiorri, & Magnavita, 2013). The
findings from the present study are also in line with previous studies that have shown
that neuroticism is associated with lower work satisfaction, higher risk of burnout, and
lower decision latitude (Alarcon et al., 2009; Grant & Langan-Fox, 2007; Judge et al.,
2002; Sutin & Costa, 2010). In addition, the current results support previous findings on
temperament traits conceptually similar to neuroticism predicting higher demands,
higher effort, lower control, lower rewards, and consequently higher job strain and
higher effort-reward imbalance (Hintsa et al., 2010a; Hintsanen et al., 2011). Individuals
high on the neuroticism scale might experience their work characteristics as more
negative and their decision latitude as limited due to their predisposition to experience
more distress. Previous studies have shown that neuroticism is correlated with stress
44
reactivity, which can influence an individual’s ability to cope with challenges and
predispose him or her to depressed mood (Felsten, 2004). Because of their worrying and
self-conscious nature, they might put in a lot of effort at work but due to their
insecurities, they might less often get actual rewards (e.g. promotion opportunities), or
they may have a subjective experience of not being esteemed at work. Neurotics may
also create a negative atmosphere at the workplace and therefore not receive the
favourable effects of social support (Goldsmith, 2007).
In addition to being associated with lower perceived job strain, high extraversion was
in the present study also associated with higher control, higher effort, and higher
reward. These results are in line with previous studies reporting associations of
extraversion with higher work satisfaction, lower risk of burnout, and higher decision
latitude (Alarcon et al., 2009; Judge et al., 2002; Sutin & Costa, 2010). Temperament
traits conceptually close to extraversion have been shown to predict higher control,
higher reward, and lower job strain, which is in line with the results of the current study
(Hintsa et al., 2010a; Hintsanen et al., 2011). Extraverted individuals are described as
sociable and friendly (McCrae & Costa, 1987) and might thereby participate in creating
a friendly and supportive working environment, which in turn reduces stress
(Goldsmith, 2007). In certain jobs extraversion can be a preferred characteristic and
employers might therefore reward this behaviour with decision latitude, promotion
possibilities, esteem, and job security. There may also be a halo-effect, that is,
employers might reward employees for the behaviour the trait induces, although it in
fact might not have any effect on work performance. In the present study, extraversion
did not show an association with effort-reward imbalance, which might indicate that the
characteristics of an extraverted individual are more important for perceiving a balance
between job demands and control than for perceiving reciprocity in efforts spent and
rewards received.
Openness was not associated with effort-reward imbalance in this study but was
associated with higher effort, higher demands, higher control and lower job strain.
These results are similar to previous findings showing that openness is associated with
higher decision latitude (Sutin & Costa, 2010). Being creative and independent might
induce a sense of control over the situation in persons high on the openness scale. A
characteristic of individuals high on the openness scale is that others view them as
45
intelligent (McCrae & Costa, 1987). They might, therefore, be given positions with
greater demands and they may actively seek jobs with higher demands to fulfil their
creativeness.
Conscientiousness was in the present study found to be related to higher reward,
higher control, and to lower job strain. Furthermore, conscientiousness was associated
with higher effort in women but with lower effort-reward imbalance only in men. These
results are in line with previous studies linking conscientiousness to high work
satisfaction and greater decision latitude (Judge et al., 2002; Sutin & Costa, 2010).
Conscientiousness is described as a tendency to be well-organized and persevering
(McCrae & Costa, 1987) and these might be characteristics that enhance the feeling of
control and skill discretion. Because conscientiousness has been shown to be associated
with increased performance at the workplace (Barrick & Mount, 1991) employers might
reward well-performing conscientious persons in various ways. In the current study,
high conscientiousness was associated with lower effort-reward imbalance only in men,
and the associations between conscientiousness and rewards were stronger in men as
compared to women, which suggests that women might not be as generously rewarded
for being conscientious as men might be. However, as rewards were assessed with self-
reports, it remains uncertain how closely our measures reflect the actual level of
rewards. The results might also suggest that conscientiousness influences women and
men differently and affects different aspects of work.
Agreeableness was associated with job strain and its components so that individuals
high in agreeableness had low demands, high control and consequently low perceived
job strain. The associations turned to non-significant when the other traits were
controlled for which might indicate that the association between agreeableness and job
strain is a reflection of the influence of the other traits on this association. Furthermore,
high agreeableness was associated with lower effort-reward imbalance, lower effort, and
higher reward. Previous studies have shown that agreeableness is associated with high
work satisfaction and lower risk of burnout (Alarcon et al., 2009; Judge et al., 2002).
The present results are in accordance with these findings. Agreeable individuals are
characterized by flexibility and sympathy (McCrae & Costa, 1987) and might therefore
not experience the demands of the work as being as restrictive as those who are low in
agreeableness. Agreeable individuals might not experience putting in a lot of effort as
46
they might enjoy their duties and, on the other hand, experience that they receive a lot of
positive feedback and rewards in turn. Employers might also reward an agreeable
individual for being flexible.
5.1.2 Bi-directional associations of cynicism and job strain
The results on the bi-directional association between cynicism and job strain showed
that high job strain at baseline was associated with higher baseline adjusted cynicism six
years later. This effect was further examined in the additional mediation analysis, which
revealed that cynicism mediated the relationship between job strain and depression. The
hypothesis on the bi-directionality was thus not supported by the present study.
However, the results bring new information on the relationship between the
psychosocial working environment and personality.
In the current study, high job strain was related to an increase in cynicism. This result
is in line with previous studies on burnout—a concept characterized by cynicism—
showing associations of high workload and high demands with burnout (Greenglass et
al., 2001; Lindblom et al., 2006; Schaufeli et al., 1996). The result can also be explained
by the social context model, which hypothesizes that cynicism is, in part, a result of an
adverse environment and psychosocial stress (Taylor et al., 1997). Indeed, the
development of cynicism may be determined by both genes and environmental factors
(Hakulinen et al., 2012; Merjonen et al., 2011; Rebollo & Boomsma, 2006), including
early childhood experiences, low family socioeconomic status (SES), parental Type A
behaviour, and Type A behaviour in childhood (Keltikangas-Järvinen & Heinonen,
2003). Although cynicism is considered to be a relatively stable trait in adulthood, mean
level change over time is possible (Hakulinen et al., 2014). Experiencing having little
control over a highly demanding job not only induces perceiving work characteristics as
stressful but might also elicit antagonistic and mistrustful thoughts and feelings in
employees about their jobs.
Cynicism is a personality construct that is closely related to many concepts of trait
psychology, such as neuroticism and agreeableness (Costa et al., 1986; Watson & Clark,
1992). Previous research has shown that job strain is associated with temperament and
personality traits conceptually similar to cynicism (Hintsa et al., 2010a; Hintsa, et al.,
2010b; Hintsanen et al., 2011). The direction of the association has been from
47
personality to job strain, not the other way round (Sutin & Costa, 2010). In the present
study, cynicism was not associated with change in job strain over six years. While the
Five-Factor Model traits are thought of as major broad traits, cynicism is described as a
narrower facet or aspect of broader traits. According to the bandwidth-fidelity literature
(Ashton, 1998), narrow traits tend to relate more strongly to narrow outcomes. The
narrow nature of cynicism might make it harder to detect prospective associations with
broad constructs like job strain.
Job strain has in several previous studies been shown to predict depression later in
life (Bonde, 2008; Virtanen et al., 2015), and cynicism has also been linked to
depressive symptoms (Heponiemi et al., 2006; Nabi et al., 2010). Additional analyses
were performed to explore whether the findings on the association between job strain
and cynicism could be explained by cynicism mediating the association between job
strain and depression. The results showed that cynicism mediated a considerable
amount (21.5%) of the effect of job strain on depression. High job strain might increase
cynical attitudes and mistrustful feelings towards others, which in turn might increase
depressive mood. The results might also be explained by previous research on the
reporting bias that affects the relationship between job strain and depression, when
measured with self-reports (Kolstad et al., 2011). The present results might indicate that
cynicism inflates the perceived relationship between job strain and depression and
should therefore be taken into account when measuring this association.
5.1.3 Longitudinal measurement invariance of the effort-reward imbalance
scales
The results on the longitudinal measurement invariance of the effort-reward imbalance
scales showed that both the effort and reward scales of the effort-reward imbalance
model achieved strict measurement invariance. Thus, the hypothesis on the invariance
of the effort-reward imbalance scales was supported, and this indicates that the effort
and reward scales measure the same latent variables over time. Furthermore, it was
found that high effort, low reward, and high effort-reward imbalance were associated
with higher risk for general stress 5 years later. Therefore, the results indicate that
effort-reward imbalance and its components have adequate criterion validity, which
supports our hypothesis. Moreover, based on the analyses that controlled for the
48
baseline general stress, high effort seem to be a valid prospective risk indicator for
higher general stress five years later. This goes beyond the scope of what is usually
demanded as evidence for adequate criterion validity. These results are in accordance
with the effort-reward imbalance model, which states that effort is spent as a part of a
contract, where sufficient rewards are expected in return (Siegrist et al., 2004). If the
rewards do not match the effort, this lack of reciprocity is considered particularly
stressful (Siegrist, 1996). Current findings also support previous studies that have
shown associations of effort-reward imbalance with decreased health and wellbeing
(Feldt et al., 2013; Hintsanen et al., 2007; van Vegchel et al., 2005).
5.2 Methodological considerations
This is, to our knowledge, the first study to examine the role of Five-Factor Model
personality traits and cynicism in perceptions of work characteristics. The sample used
in this study was fairly large, population-based, and consisted of varying occupations,
thereby being representative of the Finnish working-age population. In addition, the
measures used in this study have been validated in several studies, which increase the
comparability of our results to other studies. Furthermore, using cross-lagged structural
equation modelling when examining cynicism and job strain, allowed for examination
of the bi-directional associations over time and testing for causality of the constructs.
The associations of neuroticism and openness with demands, reported in the current
study, have not been found in other studies. The present study also reported an
association between agreeableness and job control which has not been found previously.
The discrepancies in results between previous studies and the present study might be
explained by sample size, or other sample related differences such as age differences
(the participants in the present study were somewhat younger) and differences in
measures of job demands and job control (for example, previous studies only used
measures other than the Job Content Questionnaire which is the most commonly used
measure of job strain).
The results from the present study gave support for an additive interaction between
job demands and job control rather than a multiplicative one. This means that demand
and control affect job strain separately, i.e. to maximally decrease job strain one should
49
increase control and decrease demand. Several formulations of job strain have been
accepted in job strain literature (Landsbergis et al., 1994; Schnall & Landsbergis, 1994)
and it has been stated that the implications for job redesign are similar for additive and
interactive effects (Karasek, 1989). In addition, the linear term has been shown to be the
best predictor of stress and health outcomes (Courvoisier & Perneger, 2010). Besides
the significant associations between Five-Factor Model traits and the linear term
formulation, we also found support for interaction effects in the quotient term
formulation. Replicating the results with several formulations of job strain reveals the
robustness of our results. The multiplicative interaction term was only significant for
one trait in the separately adjusted model and for two traits in the mutually adjusted
model. Other studies have also reported non-significant findings for multiplicative
interaction, while showing significant main effects (de Lange et al., 2003; Hintsa et al.,
2008; Hintsanen et al., 2005). This is, in part, due to the general nature of the demands
and control scales used in our study. In order to detect interaction effects, more
specificity in the scales is needed (de Jonge, van Vegchel, Shimazu, Schaufeli, &
Dormann, 2010).
In general, the personality traits of the Five-Factor Model explained a considerable
amount of the variance in the work characteristics. The highest explained variance was
obtained for neuroticism, which explained 17.1% of the variance in effort-reward
imbalance in women and 12% in men. In addition, neuroticism explained 15.9% of the
variance in job strain. The other traits explained less than 10% of the variance in the
work characteristics. After mutually adjusting for personality traits, the highest
explained variance was found for neuroticism in relation to job strain (7.9%). The rest
of the traits individually explained less than 3%. The complete Five-Factor personality
model explained 16.2% in the linear term job strain formulation.
In order to examine the association of a trait while simultaneously taking into account
the other personality traits, the traits were entered into the analysis simultaneously when
examining the association between the Five-Factor traits and job strain in Study II. This
way, the association between a specific trait and an outcome measure was not affected
by the other traits. The possibility of response bias due to the characteristics of
neuroticism, which has been suggested in previous research (Costa Jr. & McCrae,
1990), is controlled for using this method. However, it has been concluded that although
50
this bias exists and should be taken into account when interpreting the results, the
substantive effects of the trait also exist alongside it (Costa Jr. & McCrae, 1990; Oliver
et al., 2010; Stansfeld, 2002). Personality includes, however, a combination of all traits
and therefore the effect of a single trait on the perception of stress may not reflect the
reality. Nevertheless, examining single traits provides us with information on the nature
of the trait’s effect on job strain, which is not available when examining the whole
personality model.
Some limitations should be taken into account when interpreting our results. First, in
our study both personality traits and work characteristics were assessed with self-
reports, which constitutes a risk for common method bias. However, it has recently been
argued that common method bias is not automatically a source of bias in research that
uses self-reports (Conway & Lance, 2010) and that its effect, when it exists in
organizational research, is often rather small (Spector, 2006) and may actually decrease
the associations, not amplify them (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).
Furthermore, controlling for the baseline level of cynicism/job strain in study III,
controls for the common method bias. Second, Studies I and II were cross-sectional and
therefore no conclusions about cause-and-effect relationships or temporal precedence
can be made. This raises the question about the direction of the association. According
to the corresponsive principle of personality development, personality traits predict
particular workplace conditions and these conditions drive normative personality
development (Roberts et al., 2003). However, a recent study on the Five-Factor traits
and work characteristics has shown that the direction is one-directional, from Five-
Factor personality to occupational experiences rather than the other way around (Sutin
& Costa, 2010). In addition, results from longitudinal, population-based studies on
temperament-based personality traits and stressful work characteristics (Hintsa et al.,
2010a; Hintsanen et al., 2011) suggest that personality traits may indeed predict
perceptions of stressful work.
Third, the participants had higher educational level and occupational status in addition
to having lower scores on neuroticism. One might therefore conclude that neurotics
might have a tendency of dropping out of studies like these. These tendencies might
affect the associations so that it appears that neuroticism has weaker association with
the outcome measure than it has in reality, for example neurotics a weaker association
51
with higher perceived job strain. However, there were no differences between the
included and excluded on job strain or effort-reward imbalance which makes the
probability of our sample being highly selective low. Nevertheless our results may
reflect individuals with lower neuroticism and higher socioeconomic status slightly
better than individuals with higher neuroticism and lower socioeconomic status. In
addition, because our sample corresponds to the vast majority of the Finnish population,
our results may not be directly generalizable to other ethnic or cultural groups.
Fourth, gender differences were examined only in Study I. We decided to test the
genders separately in Study I due to significant gender interaction between some of the
study variables. Therefore the research framework differed in Studies I, II and III, which
makes the comparability of the results—in regard to gender differences—difficult. In
future studies it would be appropriate to base decisions about examining men and
women separately on theory rather than on methodology.
Fifth, the current study only included subjective measures of job strain and effort-
reward imbalance and in future studies it would be beneficial to include subjective as
well as objective measures of work characteristics, as proposed in recent literature
(Haeusser et al., 2010). It might also be beneficial for future studies to use the iso-strain
model, which also measures social support. The extended model might be able to better
mirror the characteristics of personality traits such as extraversion and agreeableness
that are thought to increase social support. It would also be appropriate for future
studies to use more narrow measures of personality traits, e.g. include the facets of the
Five-Factor Model, because they, according to the bandwidth-fidelity literature (Ashton,
1998), tend to relate more strongly to narrow outcomes like demand and control or
effort and reward.
Sixth, the Cronbach’s alphas for the job demands scale were 0.61 and 0.63, which can
be considered at the lower range of acceptable reliability estimates. However, despite
the low alpha, the mean inter-item correlation showed satisfying internal consistency for
the scale. In addition, the low alphas may reflect the small number of items in the
demand scale. Furthermore, possible job changes during the follow-up were not
inquired in our study. In future studies, measuring change in work characteristics
systematically in several time points would make the interpretations of the results more
reliable.
52
Seventh, in Study IV the RMSEA and CFI values of the invariance models for the
first-order and the second-order reward scales, showed adequate fit, rather than good fit.
However, we were able to obtain strict measurement invariance in all scales—effort,
first-order reward, and second-order reward—which indicates that the constructs are
indeed measurement invariant over time. In addition, the outcome variable for assessing
criterion validity in Study IV was measured with only one general stress item.
Nevertheless, the single-item measure of stress has been validated and shown to be a
sensitive indicator of well-being at work (Elo et al., 2003). In future studies it would be
important to also examine the longitudinal measurement invariance of the job strain
scales as they are also widely used in longitudinal and intervention studies.
5.3 Theoretical implications
The concept of individual differences in perceptions of the work environment is not a
novel one. There was a shift towards a more service-orientated and knowledge-based
economy in the 1960’s. It became important to place a human actor in the field of stress
research—not only promote physical health but also well-being of the working mind
(Väänänen et al., 2012). Lazarus and colleagues (1966; 1984) turned the focus from
structural factors at work towards individual sense-making in the stress process.
Although the focus was shifted towards the individual, the models that were developed
to conceptualize and measure work characteristics (the job strain model and the effort-
reward imbalance model) did not fully take into account that there are individual
differences in stress reactions, resilience or recovery. Instead, they were developed to
capture the work environment and depict strenuous working conditions, regardless of
the individual experiencing it. When measured with self-reports, job strain and effort-
reward imbalance cannot be completely objective and independent of individual
dispositions, such as Five-Factor personality traits. The present study adds to the work
stress literature by showing that perceptions of job strain and effort-reward imbalance at
work are susceptible to personality traits of the Five-Factor Model and that job strain
increases cynicism.
At the general level, the results can be explained by a variety of causes. According to
the differential exposure model, individuals with different personality traits place
themselves in different situations and are therefore exposed to a varying level of
53
stressors (Bolger & Zuckerman, 1995). Neurotics might choose occupations where they
are exposed to more stressors than extraverts. In addition, the selection hypothesis states
that exposure to adverse circumstances during life is not necessarily random; cynicism
has been linked to greater risk of depression (Heponiemi et al., 2006; Nabi et al., 2010)
which might also be related to higher exposure to stressors (Bonde, 2008; Virtanen et
al., 2015). Furthermore, personality is thought to affect the appraisal process;
personality influences what is perceived as stressful and what is not (Code & Langan-
Fox, 2001). Agreeable individuals might, due to their flexible nature, not experience the
demands of the work as being as strenuous as those who are low in agreeableness.
Personality also affects stress reactivity (Felsten, 2004) and thereby determines how
strongly and in what way an individual responds to stress. In terms of the differential
reactivity model (Bolger & Zuckerman, 1995), neurotics may respond differently to a
sudden increase in work load than non-neurotics. According to the psychosocial
vulnerability model, cynical individuals are more vulnerable to adverse conditions and
benefit less from psychosocial resources than non-cynical (Kivimäki et al., 2003; Smith,
1994). Indeed, cynicism has been found to be associated with more conflicts and less
social support as well as less psychophysiological benefit from social support, when
available (Lepore, 1995; Smith, Glazer, Ruiz, & Gallo, 2004). Furthermore, personality
influences what coping methods an individual uses in order to alleviate or manage stress
(e.g. Connor-Smith & Flachsbart, 2007). Conscientious individuals tend to use positive
cognitive appraisal and adaptive coping methods (Penley & Tomaka, 2002) and these
methods might help buffer the stressor.
According to the social context model, adverse conditions are an antecedent of
cynical hostility (Taylor et al., 1997). Stressful circumstances, such as putting in effort
but not receiving adequate rewards are hypothesized to lead to cynical perceptions and
inadequate coping (Taylor et al., 1997). Indeed, studies on burnout as a result of high
workload, give support to this model; burnout is characterized by cynicism and studies
show that an employee’s attitudes and behaviour change in a negative way as a result of
adverse working conditions (Greenglass et al., 2001; Lindblom et al., 2006; Schaufeli et
al., 1996). This is also in accordance with the outcome model, which suggests that
stressors influence personality, not the other way around (Kivimäki, 1996).
54
Some gender differences were found in the association between personality and work
characteristics. Low conscientiousness was associated with higher effort-reward
imbalance only in men. This suggests that conscientiousness has different roles in the
working environment for men and women. Conscientiousness seems to be more
important for men when perceiving a balance between efforts spent and rewards
received. It may also be that conscientiousness steers men and women into different
working environments and they are therefore exposed to different stressors.
It has been debated whether the job strain model and the effort-reward imbalance
model complement each other or if they actually depict the same work characteristics
(Calnan et al., 2004; Tsutsumi & Kawakami, 2004). The results of the present study
show that in terms of Five-Factor personality, the two work stress models differ to some
extent. Neuroticism and agreeableness play an important role in both models, whereas
extraversion, openness, and conscientiousness seem to be important in measuring job
strain caused by high demands combined with low control while not being important for
measuring the reciprocal nature of efforts spent and rewards received at work.
5.4 Conclusions and practical implications
The results of this study show that Five-Factor model personality traits are associated
with perceptions of work characteristics. In addition, job strain prospectively predicts
higher cynicism. Establishing the associations between an employee’s personality and
perceptions of the working environment is important in order to recognize the
predisposing and protective factors that influence vulnerability to work stress. In
addition, recognizing the effect of work on cynical attitudes is also important
information for occupational health services. This knowledge can then be utilized to
create preventive and intervention programs for reducing stress and restructuring the
working environment.
Being aware of individual dispositions influencing the well-being at work also
benefits the organization, because disregarding individual differences in perceptions of
the work environment can lead to misinterpretations, for example employers might
misinterpret the stressfulness of the work as different individuals react to and cope with
stressors differently. These misinterpretations can lead to expensive and unnecessary
55
actions that are aimed at improving the working conditions, whereas a more effective
target would sometimes be the person himself.
The present study also showed that the effort-reward imbalance scales are
measurement invariant over time and are therefore reliable measures for change in
perceptions of work characteristics. Due to the wide use of effort-reward imbalance
model to study and depict stressful working conditions, establishing measurement
invariance of the scales has implications for a large amount of longitudinal
organizational studies. In addition, knowing that changes in the scores reflect true
changes in perceptions of stressful work characteristics enables occupational health
services to target stress management and intervention appropriately.
In order to make valid conclusions about the effect of the psychosocial work
environment on employee health and wellbeing, occupational health services and
occupational psychologists should be able to rely on the measures used to conceptualize
work characteristics. Measuring work characteristics—job strain and effort-reward
imbalance—with self-reports makes them susceptible to individual dispositions such as
personality. However, automatically controlling for personality traits in research is not
encouraged, because of the information that will be lost as a result (Spector, Zapf, Chen,
& Frese, 2000). Perceptions of the environment are dependent of the person perceiving
it, and simplifying the environment to merely objective characteristics might bring
information that can help in organizing work at the organizational level, but it will not
tell occupational health services about the wellbeing of the individuals. Personality
traits should instead be used as moderators, in order to capture the effect personality has
on the association between work characteristics and well-being. In addition, future
studies should examine the interaction between personality and occupation in
explaining well-being at work—information on person-job fit could then be used in
recruiting, in stress interventions and in organizational restructuring to fit the person to
the job.
The structural aspects of the work environment are undeniably important in the
experience of stress at work, but our study suggests that personality should also be taken
into account. In addition, the working environment can have far reaching consequences
on an individual’s attitudes, which in turn can influence health and wellbeing.
Researchers and occupational health professionals should be aware that when work
56
characteristics are measured with the job strain model and the effort-reward imbalance
model, the results reflect both the structural work environment and the individual
differences that emerge through personality.
57
6 References
Åkerblom, H. K., Uhari, M., Pesonen, E., Dahl, M., Kaprio, E. A., Nuutinen, E. M., . . . Viikari, J. (1991).
Cardiovascular risk in young Finns. Annals of Medicine, 23(1), 35–39.
doi:10.3109/07853899109147928
Alarcon, G., Eschleman, K. J., & Bowling, N. A. (2009). Relationships between personality variables and
burnout: A meta-analysis. Work and Stress, 23(3), 244–263. doi:10.1080/02678370903282600
Anusic, I., Lucas, R. E., & Donnellan, M. B. (2012). Cross-sectional age differences in personality:
Evidence from nationally representative samples from Switzerland and the United States. Journal of
Research in Personality, 46(1), 116–120. doi:10.1016/j.jrp.2011.11.002
Ashton, M. C. (1998). Personality and job performance: The importance of narrow traits. Journal of
Organizational Behavior, 19(3), 289–303. doi:10.1002/(SICI)1099-1379(199805)19:3<289::AID-
JOB841>3.3.CO;2-3
Backe, E., Seidler, A., Latza, U., Rossnagel, K., & Schumann, B. (2012). The role of psychosocial stress
at work for the development of cardiovascular diseases: A systematic review. International Archives
of Occupational and Environmental Health, 85(1), 67–79. doi:10.1007/s00420-011-0643-6
Barrick, M. R., & Mount, M. K. (1991). The big 5 personality dimensions and job-performance - a meta-
analysis. Personnel Psychology, 44(1), 1–26. doi:10.1111/j.1744-6570.1991.tb00688.x
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the beck depression inventory-II. San
Antonio, TX: Psychological Corporation.
Belkic, K. L., Landsbergis, P. A., Schnall, P. L., & Baker, D. (2004). Is job strain a major source of
cardiovascular disease risk? Scandinavian Journal of Work Environment & Health, 30(2), 85–128.
Bolger, N., & Zuckerman, A. (1995). A framework for studying personality in the stress process. Journal
of Personality and Social Psychology, 69(5), 890–902. doi:10.1037//0022-3514.69.5.890
Bonde, J. P. E. (2008). Psychosocial factors at work and risk of depression: A systematic review of the
epidemiological evidence. Occupational and Environmental Medicine, 65(7), 438–445.
doi:10.1136/oem.2007.038430
Bourbonnais, R., Brisson, C., & Vezina, M. (2011). Long-term effects of an intervention on psychosocial
work factors among healthcare professionals in a hospital setting. Occupational and Environmental
Medicine, 68(7), 479–486. doi:10.1136/oem.2010.055202
58
Brunner, E., Kivimäki, M., Siegrist, J., Theorell, T., Luukkonen, R., Riihimäki, H., . . . Leino-Arjas, P.
(2004). Is the effect of work stress on cardiovascular mortality confounded by socioeconomic
factors in the Valmet study? Journal of Epidemiology and Community Health, 58(12), 1019–1020.
doi:10.1136/jech.2003.016881
Calnan, M., Wadsworth, E., May, M., Smith, A., & Wainwright, D. (2004). Job strain, effort-reward
imbalance, and stress at work: Competing or complementary models? Scandinavian Journal of
Public Health, 32(2), 84–93. doi:10.1080/14034940310001668
Cannon, W. B. (1915). Bodily changes in pain, hunger, fear and rage, an account of recent researches
into the function of emotional excitement. New York: D. Appleton and Company.
Caspi, A., Wright, B. R. E., Moffitt, T. E., & Silva, P. A. (1998). Early failure in the labor market:
Childhood and adolescent predictors of unemployment in the transition to adulthood. American
Sociological Review, 63(3) doi:10.2307/2657557
Chandola, T., Heraclides, A., & Kumari, M. (2010). Psychophysiological biomarkers of workplace
stressors. Neuroscience and Biobehavioral Reviews, 35(1), 51–57.
doi:10.1016/j.neubiorev.2009.11.005
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit indexes for testing measurement
invariance. Structural Equation Modeling, 9(2), 233–255. doi:10.1207/S15328007SEM0902_5
Chida, Y., & Steptoe, A. (2009). The association of anger and hostility with future coronary heart disease
A meta-analytic review of prospective evidence. Journal of the American College of Cardiology,
53(11) doi:10.1016/j.jacc.2008.11.044
Code, S., & Langan-Fox, J. (2001). Motivation, cognitions and traits: Predicting occupational health,
well-being and performance. Stress and Health, 17(3), 159–174. doi:10.1002/smi.897
Comrey, A. L. (1957). A factor-analysis of items on the mmpi depression scale. Educational and
Psychological Measurement, 17(4), 578–585.
Comrey, A. L. (1958). A factor-analysis of items on the mmpi paranoia scale. Educational and
Psychological Measurement, 18(1), 99–107. doi:10.1177/001316445801800109
Connor-Smith, J. K., & Flachsbart, C. (2007). Relations between personality and coping: A meta-
analysis. Journal of Personality and Social Psychology, 93(6), 1080–1107. doi:10.1037/0022-
3514.93.6.1080
59
Conway, J. M., & Lance, C. E. (2010). What reviewers should expect from authors regarding common
method bias in organizational research. Journal of Business and Psychology, 25(3), 325–334.
doi:10.1007/s10869-010-9181-6
Costa Jr., P. T., & McCrae, R. R. (1990). Personality: Another 'hidden factor' is stress research.
Psychological Inquiry, 1(1), 22.
Costa, P. T., Busch, C. M., Zonderman, A. B., & Mccrae, R. R. (1986). Correlations of mmpi factor
scales with measures of the 5-factor model of personality. Journal of Personality Assessment, 50(4),
640–650. doi:10.1207/s15327752jpa5004_10
Costa, P. T., & Mccrae, R. R. (1985). The NEO personality inventory manual. Odessa, FL: Psychological
Assessment Resources.
Costa, P. T., & Mccrae, R. R. (1989). The NEO-PI/NEO-FFI manual supplement. Odessa, FL:
Psychological Assessment Records.
Courvoisier, D. S., & Perneger, T. V. (2010). Validation of alternative formulations of job strain. Journal
of Occupational Health, 52(1), 5–13.
de Jonge, J., van der Linden, S., Schaufeli, W., Peter, R., & Siegrist, J. (2008). Factorial invariance and
stability of the effort-reward imbalance scales: A longitudinal analysis of two samples with different
time lags. International Journal of Behavioral Medicine, 15(1), 62–72.
doi:10.1080/10705500701783959
de Jonge, J., van Vegchel, N., Shimazu, A., Schaufeli, W., & Dormann, C. (2010). A longitudinal test of
the demand-control model using specific job demands and specific job control. International
Journal of Behavioral Medicine, 17(2), 125–133. doi:10.1007/s12529-010-9081-1
De Lange, A. H., Taris, T. W., Jansen, P., Kompier, M. A. J., Houtman, I. L. D., & Bongers, P. M.
(2010). On the relationships among work characteristics and learning-related behavior: Does age
matter? Journal of Organizational Behavior, 31(7), 925–950. doi:10.1002/job.649
de Lange, A. H., Taris, T. W., Kompier, M. A. J., Houtman, I. L. D., & Bongers, P. M. (2003). "The very
best of the millennium": Longitudinal research and the demand-control-(support) model. Journal of
Occupational Health Psychology, 8(4), 282–305.
60
Ebstrup, J. F., Eplov, L. F., Pisinger, C., & Jorgensen, T. (2011). Association between the five factor
personality traits and perceived stress: Is the effect mediated by general self-efficacy? Anxiety Stress
and Coping, 24(4), 407–419. doi:10.1080/10615806.2010.540012
Elo, A., Leppänen, A., & Jahkola, A. (2003). Validity of a single-item measure of stress symptoms.
Scandinavian Journal of Work Environment & Health, 29(6), 444–451.
Elo, A., Leppänen, A., Lindström, K., & Ropponen, T. (1992). Occupational stress questionnaire: User's
instruction (reviews 19). Helsinki: Finnish Institute of Occupational Health.
Elovainio, M., Kivimäki, M., Kortteinen, M., & Tuomikoski, H. (2001). Socioeconomic status, hostility
and health. Personality and Individual Differences, 31(3), 303–315. doi:10.1016/S0191-
8869(00)00137-9
Elovainio, M., Kivimäki, M., Vahtera, J., Virtanen, M., & Keltikangas-Järvinen, L. (2003). Personality as
a moderator in the relations between perceptions of organizational justice and sickness absence.
Journal of Vocational Behavior, 63(3), 379–395. doi:10.1016/S0001-8791(02)00058-1
Elovainio, M., Salo, P., Jokela, M., Heponiemi, T., Linna, A., Virtanen, M., . . . Vahtera, J. (2013).
Psychosocial factors and well-being among Finnish GPs and specialists: A 10-year follow-up.
Occupational and Environmental Medicine, 70(4), 246–251. doi:10.1136/oemed-2012-100996
EU-OSHA. (2009). European agency for safety and health at work: OSH in figures: Stress at work - facts
and figures. Luxembourg: Office for Official Publications of the European Communities.
EU-OSHA. (2013). European agency for safety and health at work: European opinion poll on
occupational safety and health. Retrieved from https://osha.europa.eu/en/safety-health-in-figures/eu-
poll-press-kit-2013.pdf
Feldt, T., Huhtala, M., Kinnunen, U., Hyvonen, K., Mäkikangas, A., & Sonnentag, S. (2013). Long-term
patterns of effort-reward imbalance and over-commitment: Investigating occupational well-being
and recovery experiences as outcomes. Work and Stress, 27(1), 64–87.
doi:10.1080/02678373.2013.765670
Felsten, G. (2004). Stress reactivity and vulnerability to depressed mood in college students. Personality
and Individual Differences, 36(4), 789–800. doi:10.1016/S0191-8869(03)00152-1
61
Garbarino, S., Cuomo, G., Chiorri, C., & Magnavita, N. (2013). Association of work-related stress with
mental health problems in a special police force unit. BMJ Open, 3(7), e002791.
doi:10.1136/bmjopen-2013-002791
Godin, I., Kittel, F., Coppieters, Y., & Siegrist, J. (2005). A prospective study of cumulative job stress in
relation to mental health. BMC Public Health, 5, 67. doi:10.1186/1471-2458-5-67
Goldsmith, E. (2007). Stress, fatigue, and social support in the work and family context. Journal of Loss
& Trauma, 12(2), 155–169. doi:10.1080/15434610600854228
Grant, S., & Langan-Fox, J. (2007). Personality and the occupational stressor-strain relationship: The role
of the big five. Journal of Occupational Health Psychology, 12(1), 20–33. doi:10.1037/1076-
8998.12.1.20
Greenglass, E. R., Burke, R. J., & Moore, K. A. (2003). Reactions to increased workload: Effects on
professional efficacy of nurses. Applied Psychology-an International Review-Psychologie
Appliquee-Revue Internationale, 52(4), 580–597. doi:10.1111/1464-0597.00152
Greenglass, E., Burke, R., & Fiksenbaum, L. (2001). Workload and burnout in nurses. Journal of
Community & Applied Social Psychology, 11(3), 211–215. doi:10.1002/casp.614
Greenglass, E., & Julkunen, J. (1989). Construct-validity and sex-differences in cook-medley hostility.
Personality and Individual Differences, 10(2), 209–218. doi:10.1016/0191-8869(89)90206-7
Haeusser, J. A., Mojzisch, A., Niesel, M., & Schulz-Hardt, S. (2010). Ten years on: A review of recent
research on the job demand-control (-support) model and psychological well-being. Work and
Stress, 24(1), 1–35. doi:10.1080/02678371003683747
Hakulinen, C., Jokela, M., Hintsanen, M., Merjonen, P., Pulkki-Råback, L., Seppälä, I., . . . Keltikangas-
Järvinen, L. (2012). Serotonin receptor 1B genotype and hostility, anger and aggressive behavior
through the lifespan: The young Finns study. Journal of Behavioral Medicine, , 1–8.
doi:10.1007/s10865-012-9452-y
Hakulinen, C., Jokela, M., Hintsanen, M., Pulkki-Råback, L., Elovainio, M., Hintsa, T., . . . Keltikangas-
Järvinen, L. (2013). Hostility and unemployment: A two-way relationship? Journal of Vocational
Behavior, 83(2), 153–160. doi:10.1016/j.jvb.2013.04.003
62
Hakulinen, C., Jokela, M., Keltikangas-Järvinen, L., Merjonen, P., Raitakari, O. T., & Hintsanen, M.
(2014). Longitudinal measurement invariance, stability and change of anger and cynicism. Journal
of Behavioral Medicine, 37(3), 434–44. doi:10.1007/s10865-013-9501-1
Heponiemi, T., Elovainio, M., Kivimäki, M., Pulkki, L., Puttonen, S., & Keltikangas-Järvinen, L. (2006).
The longitudinal effects of social support and hostility on depressive tendencies. Social Science &
Medicine, 63(5), 1374–1382. doi:10.1016/j.socscimed.2006.03.036
Hintsa, T., Kivimäki, M., Elovainio, M., Vahtera, J., Hintsanen, M., Viikari, J. S. A., . . . Keltikangas-
Järvinen, L. (2008). Is the association between job strain and carotid intima-media thickness
attributable to pre-employment environmental and dispositional factors? the cardiovascular risk in
young Finns study. Occupational and Environmental Medicine, 65(10), 676–682.
doi:10.1136/oem.2007.037622
Hintsa, T., Hintsanen, M., Jokela, M., Elovainio, M., Raitakari, O., & Keltikangas-Järvinen, L. (2010a).
The influence of temperament on long-term job strain and its components: The cardiovascular risk
in young Finns study. Personality and Individual Differences, 49(7), 700–705.
doi:10.1016/j.paid.2010.06.007
Hintsa, T., Hintsanen, M., Jokela, M., Pulkki-Råback, L., & Keltikangas-Järvinen, L. (2010b). Divergent
influence of different type A dimensions on job strain and effort-reward imbalance. Journal of
Occupational and Environmental Medicine, 52(1), 1–7. doi:10.1097/JOM.0b013e3181c559ea
Hintsanen, M., Kivimäki, M., Elovainio, M., Pulkki-Råback, L., Keskivaara, P., Juonala, M., . . .
Keltikangas-Järvinen, L. (2005). Job strain and early atherosclerosis: The cardiovascular risk in
young finns study. Psychosomatic Medicine, 67(5), 740–747.
doi:10.1097/01.psy.0000181271.04169.93
Hintsanen, M., Elovainio, M., Puttonen, S., Kivimäki, M., Koskinen, T., Raitakari, O. T., & Keltikangas-
Järvinen, L. (2007). Effort-reward imbalance, heart rate, and heart rate variability: The
cardiovascular risk in young Finns study. International Journal of Behavioral Medicine, 14(4), 202–
212.
Hintsanen, M., Hintsa, T., Widell, A., Kivimäki, M., Raitakari, O. T., & Keltikangas-Järvinen, L. (2011).
Negative emotionality, activity, and sociability temperaments predicting long-term job strain and
63
effort-reward imbalance: A 15-year prospective follow-up study. Journal of Psychosomatic
Research, 71(2), 90–96. doi:10.1016/j.jpsychores.2011.02.012
Hjemdahl, P., Rosengren, A., & Steptoe, A. (Eds.). (2011). Stress and cardiovascular disease. New York:
Springer-Verlag.
Johnson, J., & Hall, E. (1988).
Job strain, work place social support, and cardiovascular-disease - a cross-sectional study of a
random sample of the Swedish working population. American Journal of Public Health, 78(10),
1336.
Judge, T., Heller, D., & Mount, M. (2002). Five-factor model of personality and job satisfaction: A meta-
analysis. Journal of Applied Psychology, 87(3), 530–541. doi:10.1037//0021-9010.87.3.530
Karasek, R. A. (1979). Job demands, job decision latitude, and mental strain - implications for job
redesign. Administrative Science Quarterly, 24(2), 285–308. doi:10.2307/2392498
Karasek, R. A. (1985). Job content questionnaire and user's guide. revision 1.1. Los Angeles: Department
of Industrial and Systems Engineering, University of Southern Los Angeles.
Karasek, R. A. (1989). Control in the workplace and its health-related aspects. In S. L. Sauter, J. J.
Hurrell & C. L. Cooper (Eds.), Job control and work health (pp. 129–159). New York: Wiley.
Karasek, R. A., & Theorell, T. (1990). Healthy work: Stress, productivity and the reconstruction of
working life. New York: Basic Books.
Kauppinen, T., Mattila-Holappa, P., Perkiö-Mäkelä, M., Saalo, A. T.,J., Tuomivaara, S., Uuksulainen, S.,
. . . Virtanen, S. (2013). Työ ja terveys suomessa 2012. Helsinki: Työterveyslaitos.
Keltikangas-Järvinen, L., & Heinonen, K. (2003). Childhood roots of adulthood hostility: Family factors
as predictors of cognitive and affective hostility. Child Development, 74(6), 1751–1768.
doi:10.1046/j.1467-8624.2003.00636.x
Kivimäki, M. (1996). Stress and personality factors. (No. People and work, Research Reports 9).
Helsinki: Finnish Institute of Occupational Health.
Kivimäki, M., Elovainio, M., Kokko, K., Pulkkinen, L., Kortteinen, M., & Tuomikoski, H. (2003).
Hostility, unemployment and health status: Testing three theoretical models. Social Science &
Medicine, 56(10), 2139–2152. doi:10.1016/S0277-9536(02)00219-8
64
Kivimäki, M., Nyberg, S. T., Batty, G. D., Fransson, E. I., Heikkilä, K., Alfredsson, L., . . . Theorell, T.
(2012). Job strain as a risk factor for coronary heart disease: A collaborative meta-analysis of
individual participant data. Lancet, 380(9852), 1491–1497. doi:10.1016/S0140-6736(12)60994-5
Kivimäki, M., Nyberg, S. T., Fransson, E. I., Heikkilä, K., Alfredsson, L., Casini, A., . . . Ipd-Work
Consortium. (2013). Associations of job strain and lifestyle risk factors with risk of coronary artery
disease: A meta-analysis of individual participant data. Canadian Medical Association Journal,
185(9), 763–769. doi:10.1503/cmaj.121735
Kolstad, H. A., Hansen, A. M., Kaergaard, A., Thomsen, J. F., Kaerlev, L., Mikkelsen, S., . . . Bonde, J.
P. (2011). Job strain and the risk of depression: Is reporting biased? American Journal of
Epidemiology, 173(1), 94–102. doi:10.1093/aje/kwq318
Landsbergis, P. A., Schnall, P. L., Warren, K., Pickering, T. G., & Schwartz, J. E. (1994). Association
between ambulatory blood-pressure and alternative formulations of job strain. Scandinavian Journal
of Work Environment & Health, 20(5), 349–363.
Lazarus, R. (1966). Psychological stress and the coping process. New York: McGraw-Hill.
Lazarus, R. (1999). Stress and emotion. A new synthesis. Chippenham: Springer Publishing Company Inc.
Lazarus, R., & Folkman, S. (1984). Stress, appraisal, and coping. . New York: Springer.
Lepore, S. (1995). Cynicism, social support, and cardiovascular reactivity. Health Psychology, 14(3),
210–216. doi:10.1037//0278-6133.14.3.210
Limm, H., Guendel, H., Heinmueller, M., Marten-Mittag, B., Nater, U. M., Siegrist, J., & Angerer, P.
(2011). Stress management interventions in the workplace improve stress reactivity: A randomised
controlled trial. Occupational and Environmental Medicine, 68(2), 126–133.
doi:10.1136/oem.2009.054148
Lindblom, K., Linton, S., Fedeli, C., & Bryngelsson, I. (2006). Burnout in the working population:
Relations to psychosocial work factors. International Journal of Behavioral Medicine, 13(1), 51–59.
doi:10.1207/s15327558ijbm1301_7
Lovallo, W. R. (1997). Stress & health. California: Sage Publications.
MacCallum, R. C., Zhang, S. B., Preacher, K. J., & Rucker, D. D. (2002). On the practice of
dichotomization of quantitative variables. Psychological Methods, 7(1), 19–40. doi:10.1037//1082-
989X.7.1.19
65
McCrae, R. R., & Costa, P. T. (1987). Validation of the 5-factor model of personality across instruments
and observers. Journal of Personality and Social Psychology, 52(1), 81–90. doi:10.1037/0022-
3514.52.1.81
McCrae, R. R., & Costa, P. T. (2003). Personality in adulthood: Emerging lives enduring dispositions.
New York: Guilford press.
Merjonen, P., Keltikangas-Järvinen, L., Jokela, M., Seppälä, I., Lyytikäinen, L., Pulkki-Råback, L., . . .
Lehtimäki, T. (2011). Hostility in adolescents and adults: A genome-wide association study of the
young Finns. Translational Psychiatry, 1, e11. doi:10.1038/tp.2011.13
Moyle, P. (1995). The role of negative affectivity in the stress process - tests of alternative models.
Journal of Organizational Behavior, 16(6), 647–668. doi:10.1002/job.4030160705
Nabi, H., Singh-Manoux, A., Ferrie, J. E., Marmot, M. G., Melchior, M., & Kivimäki, M. (2010).
Hostility and depressive mood: Results from the Whitehall II prospective cohort study.
Psychological Medicine, 40(3), 405–413. doi:10.1017/S0033291709990432
Niedhammer, I., Tek, M. L., Starke, D., & Siegrist, J. (2004). Effort-reward imbalance model and self-
reported health: Cross-sectional and prospective findings from the GAZEL cohort. Social Science &
Medicine, 58(8), 1531–1541. doi:10.1016/S0277-9536(03)00346-0
Oliver, J. E., Mansell, A., & Jose, P. E. (2010). A longitudinal study of the role of negative affectivity on
the work stressor-strain process. International Journal of Stress Management, 17(1), 56–77.
doi:10.1037/a0017696
Penley, J. A., & Tomaka, J. (2002). Associations among the big five, emotional responses, and coping
with acute stress. Personality and Individual Differences, 32(7), 1215–1228. doi:10.1016/S0191-
8869(01)00087-3
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in
behavioral research: A critical review of the literature and recommended remedies. Journal of
Applied Psychology, 88(5), 879–903. doi:10.1037/0021-9010.88.5.879
Pulver, A., Allik, J., Pulkkinen, L., & Hämäläinen, M. (1995). A big 5 personality-inventory in 2 non-
Indo-European languages. European Journal of Personality, 9(2), 109–124.
doi:10.1002/per.2410090205
66
Raitakari, O. T., Juonala, M., Rönnemaa, T., Keltikangas-Järvinen, L., Räsänen, L., Pietikäinen, M., . . .
Viikari, J. S. A. (2008). Cohort profile: The cardiovascular risk in young Finns study. International
Journal of Epidemiology, 37(6), 1220–1226. doi:10.1093/ije/dym225
Rantanen, J., Feldt, T., Hyvönen, K., Kinnunen, U., & Mäkikangas, A. (2013). Factorial validity of the
effort-reward imbalance scale: Evidence from multi-sample and three-wave follow-up studies.
International Archives of Occupational and Environmental Health, 86(6), 645–656.
doi:10.1007/s00420-012-0798-9
Rantanen, J., Metsäpelto, R., Feldt, T., Pulkkinen, L., & Kokko, K. (2007). Long-term stability in the big
five personality traits in adulthood. Scandinavian Journal of Psychology, 48(6), 511–518.
doi:10.1111/j.1467-9450.2007.00609.x
Rebollo, I., & Boomsma, D. (2006). Genetic analysis of anger: Genetic dominance or competitive sibling
interaction. Behavior Genetics, 36(2), 216–228. doi:10.1007/s10519-005-9025-8
Roberts, B. W., Caspi, A., & Moffitt, T. E. (2003). Work experiences and personality development in
young adulthood. Journal of Personality and Social Psychology, 84(3), 582–593. doi:10.1037/0022-
3514.84.3.582
Roberts, B. W., & Mroczek, D. (2008). Personality trait change in adulthood. Current Directions in
Psychological Science, 17(1), 31–35. doi:10.1111/j.1467-8721.2008.00543.x
Schaufeli, W., Leiter, M. P., Maslach, C., & Jackson, S. E. (1996). MBI - general survey [null]. Palo Alto,
CA: Consulting Psychologists Press.
Schmitt, N., & Kuljanin, G. (2008). Measurement invariance: Review of practice and implications.
Human Resource Management Review, 18(4), 210–222. doi:10.1016/j.hrmr.2008.03.003
Schnall, P. L., & Landsbergis, P. A. (1994). Job strain and cardiovascular disease. Annual Review of
Public Health, 15, 381-411.
Selye, H. (1956). The stress of life. New York: McGraw-Hill Book Company.
Shultz, K. S., Wang, M., Crimmins, E. M., & Fisher, G. G. (2010). Age differences in the demand-control
model of work stress an examination of data from 15 European countries. Journal of Applied
Gerontology, 29(1), 21–47. doi:10.1177/0733464809334286
Siegler, I. C., Costa, P. T., Brummett, B. H., Helms, M. J., Barefoot, J. C., Williams, R. D., . . . Rimer, B.
K. (2003). Patterns of change in hostility from college to midlife in the UNC alumni heart study
67
predict high-risk status. Psychosomatic Medicine, 65(5), 738–745.
doi:10.1097/01.PSY.0000088583.25140.9C
Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational
Health Psychology, 1(1), 27–41.
Siegrist, J. (2001). A theory of occupational stress. In J. Dunham (Ed.), Stress in the workplace: Past,
present and future (pp. 52-66). London: Whurr Publishers, Ltd.
Siegrist, J., & Peter, R. (1996). Measuring effort-reward imbalance at work: Guidelines. Düsseldorf:
Heinrich Heine University.
Siegrist, J., Starke, S., Chandola, T., Godin, I., Marmot, M., Niedhammer, I., & Peter, R. (2004). The
measurement of effort-reward imbalance at work: European comparisons. Social Science &
Medicine, 58(8), 1483–1499. doi:10.1016/S0277-9536(03)00351-4
Siegrist, J., & Roedel, A. (2006). Work stress and health risk behavior. Scandinavian Journal of Work
Environment & Health, 32(6), 473–481.
Smith, T. W. (1994). Concepts and methods in the study of anger, hostility, and health. In A. W.
Siegman, & T. W. Smith (Eds.), Anger, hostility and the heart (pp. 23–42). Hillsdale, NJ.: Lawrence
Erlbaum.
Smith, T., Glazer, K., Ruiz, J., & Gallo, L. (2004). Hostility, anger, aggressiveness, and coronary heart
disease: An interpersonal perspective on personality, emotion, and health. Journal of Personality,
72(6), 1217–1270. doi:10.1111/j.1467-6494.2004.00296.x
Soto, C. J., & John, O. P. (2012). Development of big five domains and facets in adulthood: Mean-level
age trends and broadly versus narrowly acting mechanisms. Journal of Personality, 80(4), 881–914.
doi:10.1111/j.1467-6494.2011.00752.x
Specht, J., Egloff, B., & Schmukle, S. C. (2011). Stability and change of personality across the life
course: The impact of age and major life events on mean-level and rank-order stability of the big
five. Journal of Personality and Social Psychology, 101(4), 862–882. doi:10.1037/a0024950
Spector, P. E. (2006). Method variance in organizational research - truth or urban legend? Organizational
Research Methods, 9(2), 221–232. doi:10.1177/1094428105284955
68
Spector, P., Zapf, D., Chen, P., & Frese, M. (2000). Why negative affectivity should not be controlled in
job stress research: Don't throw out the baby with the bath water. Journal of Organizational
Behavior, 21(1), 79–95. doi:10.1002/(SICI)1099-1379(200002)21:1<79::AID-JOB964>3.3.CO;2-7
Stansfeld, S. (2002). Work, personality and mental health. British Journal of Psychiatry, 181, 96–98.
Stansfeld, S. A., Shipley, M. J., Head, J., & Fuhrer, R. (2012). Repeated job strain and the risk of
depression: Longitudinal analyses from the Whitehall II study. American Journal of Public Health,
102(12), 2360–2366. doi:10.2105/AJPH.2011.300589
Stansfeld, S., & Candy, B. (2006). Psychosocial work environment and mental health - a meta-analytic
review. Scandinavian Journal of Work Environment & Health, 32(6), 443–462.
Steptoe, A., & Kivimäki, M. (2012). Stress and cardiovascular disease. Nature Reviews Cardiology, 9(6),
360–370. doi:10.1038/nrcardio.2012.45
Sutin, A. R., & Costa, P. T.,Jr. (2010). Reciprocal influences of personality and job characteristics across
middle adulthood. Journal of Personality, 78(1), 257–288. doi:10.1111/j.1467-6494.2009.00615.x
Taylor, S., Repetti, R., & Seeman, T. (1997). Health psychology: What is an unhealthy environment and
how does it get under the skin? Annual Review of Psychology, 48, 411–447.
doi:10.1146/annurev.psych.48.1.411
Theorell, T. (2003). Organisational justice and health of employees: Prospective cohort study -
commentary. Occupational and Environmental Medicine, 60(1), 33–34.
Tsutsumi, A., & Kawakami, N. (2004). A review of empirical studies on the model of effort-reward
imbalance at work: Reducing occupational stress by implementing a new theory. Social Science &
Medicine, 59(11), 2335–2359. doi:10.1016/j.socscimed.2004.03.030
Väänänen, A., Anttila, E., Turtiainen, J., & Varje, P. (2012). Formulation of work stress in 1960-2000:
Analysis of scientific works from the perspective of historical sociology. Social Science &
Medicine, 75(5), 784–794. doi:10.1016/j.socscimed.2012.04.014
Vahtera, J., Kivimäki, M., Uutela, A., & Pentti, J. (2000). Hostility and ill health: Role of psychosocial
resources in two contexts of working life. Journal of Psychosomatic Research, 48(1), 89–98.
doi:10.1016/S0022-3999(99)00080-X
69
van der Doef, M., & Maes, S. (1998). The job demand control (-support) model and physical health
outcomes: A review of the strain and buffer hypothesis. Psychology & Health, 13(5), 909–936.
doi:10.1080/08870449808407440
van der Noordt, M., IJzelenberg, H., Droomers, M., & Proper, K. I. (2014). Health effects of employment:
A systematic review of prospective studies. Occupational and Environmental Medicine, 71(10),
730–736. doi:10.1136/oemed-2013-101891
van Vegchel, N., de Jonge, J., Bosma, H., & Schaufeli, W. (2005). Reviewing the effort-reward
imbalance model: Drawing up the balance of 45 empirical studies. Social Science & Medicine,
60(5), 1117–1131. doi:10.1016/j.socscimed.2004.06.043
Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurement invariance
literature: Suggestions, practices, and recommendations for organizational research. Organizational
Research Methods, 3(1), 4–70. doi:10.1177/109442810031002
Vandervoort, D. (1999). Quality of social support in mental and physical health. Current Psychology,
18(2), 205–222. doi:10.1007/s12144-999-1029-8
Virtanen, M., Ferrie, J. E., Batty, G. D., Elovainio, M., Jokela, M., Vahtera, J., . . . Kivimaki, M. (2015).
Socioeconomic and psychosocial adversity in midlife and depressive symptoms post retirement: A
21-year follow-up of the Whitehall II study. American Journal of Geriatric Psychiatry, 23(1), 99–
U129. doi:10.1016/j.jagp.2014.04.001
Watson, D., & Clark, L. A. (1992). On traits and temperament - general and specific factors of emotional
experience and their relation to the 5-factor model. Journal of Personality, 60(2), 441–476.
doi:10.1111/j.1467-6494.1992.tb00980.x