Content uploaded by Wanja Astvik
Author content
All content in this area was uploaded by Wanja Astvik on Sep 23, 2014
Content may be subject to copyright.
Journal of Social Work
13(4) 337–360
!The Author(s) 2012
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1468017311434682
jsw.sagepub.com
Article
Coping with the
imbalance between job
demands and resources:
A study of different
coping patterns and
implications for health
and quality in human
service work
Wanja Astvik and Marika Melin
Ma
¨lardalen University; Stockholm University, Sweden
Abstract
Summary: In recent decades the public welfare sector has been subjected to major
structural changes, and studies of various occupational groups within human service
work have reported increased workload and a high prevalence of work-related stress.
Using questionnaire data from a sample of human service workers within social
work, child care and elderly care, the aim of this study was to identify different pat-
terns of coping strategies to manage the imbalance between work demands and
resources, and then to investigate their impact on outcomes in employee health and
service quality.
Findings: Cluster analysis identified three strategy profiles: compensatory and quality
reducing,voice and support seeking and self-supporting, and the comparative analysis
indicated that the compensatory and quality reducing cluster may be regarded as a
risk group. Results of hierarchical regression analyses disclosed that the identified
strategies affected health outcomes as well as perceived service quality. The use of
compensatory and quality reducing strategies was negatively related to health and
quality, although work demands, resources and background characteristics were
controlled for.
Applications: The results add to the research field through the identification of
compensatory and quality reducing strategies not previously described in the coping
Corresponding author:
Wanja Astvik, Department of Psychology, Ma
¨lardalen University, PO Box 883, SE-721 23 Va
¨stera
˚s, Sweden.
Email: wanja.astvik@mdh.se
literature, as well as the risks associated with them. Applied in practice, the identified
strategy clusters might help distinguish ‘risk behaviors’ from more beneficial strategies.
The results also point toward the importance of providing organizational structures that
allow the employees to voice their opinions and critique, as well as to give and receive
social support.
Keywords
Coping, employee health, human service work, job stress, service quality, social work
Introduction
In recent decades the public welfare sector has been subjected to major structural
changes. Cutbacks and a growing emphasis on cost efficiency as well as more
process-oriented, decentralized ways of working have changed working condi-
tions for employees. The restructuring of human service organizations into
leaner organizations also appears to have brought increased work demands.
Many employees in this sector perceive their situation as being characterized
by high workload and insufficient resources to satisfactorily carry out the work
(Demerouti, Bakker, Nachreiner, & Shaufeli, 2001; Hellgren, Na
¨swall, Sverke, &
So
¨derfeldt, 2003). The workload has increased in terms of both the amount of
work and the complexity of assigned work tasks (Peeters & Le Blanc, 2001).
Several studies on the work conditions of various occupational groups within
the public welfare sector have reported high and increased workload as well as
decreased employee control (e.g. Ha
¨renstam & MOA Research Group, 2005;
Ha
¨renstam, Bejerot, Leijon, Sche
´ele, & Waldenstro
¨m, 2004; Petterson, Hertting,
Hagberg, & Theorell, 2005). Available data also indicate a stable trend of a high
prevalence of work-related stress among occupations in the public welfare sector
(European Agency for Safety and Health at Work, 2009). Social work stands out
as a particular exposed occupational group among human service professions. In
a comparison with other human service professions, social workers and especially
child welfare social workers report higher workload and more difficult demands
(Tham & Meagher, 2008). Several studies of work conditions in social work have
found that this group experience high workload as one of the main sources of
stress at work (Balloch, Pahl, & McLean, 1998; Coffey, Dugdill, & Tattersall,
2004, 2009; Collings & Murray, 1996; Huxley et al., 2005) and high work
demands seems to be associated with depression (Stanley, Manthorpe, &
White, 2007). Insufficient resources leading to unmanageable workload have
also been found to be an important reason to why people leave their occupation
in social services (Audit Commission, 2002; Healy, Meagher, & Cullin, 2009). In
Sweden, as in many other countries, problems with recruitment as well as reten-
tion of social workers have been reported. In a Swedish study (Tham, 2007) of
child welfare social workers it was found that although 54 percent of the sample
338 Journal of Social Work 13(4)
had been at their current workplace for two years or less, 48 percent intended to
leave their jobs. The most important reason for intending to leave was however
not high workload, but lack of human resource orientation in the organization,
that is, the extent to which employees were rewarded for a job well done,
felt taken care of and where the management was interested in their health
and well-being.
Another general trend in human service work and in modern working life in
general, concerns the deregulation and decentralization of responsibility, demand-
ing a greater independence and self-regulation from employees (Allvin, 2008;
Hellgren, Sverke, & Na
¨swall, 2008). This increased individual responsibility for
work, combined with high workload and insufficient resources, brings employees’
coping strategies to the fore in understanding stress and employee health in human
service work.
In this article, the aim is to identify different patterns of coping strategies
adopted by human service workers for managing the imbalance between work
demands and resources, and then to investigate what impact these coping strategy
patterns have on outcomes in employee health and service quality. The hypothesis
tested in this study is that people confronted with situations of imbalance between
demands and resources develop different, more or less beneficial and constructive
ways to cope with these situations, and depending on what kind of strategies they
adopt, their health and the service quality of their work is affected in a positive or
negative way. In order to investigate this, the first objective of the study was to
identify patterns of coping strategies connected to excessive work demands, and
second, to distinguish differences between these coping clusters (regarding
background characteristics, working conditions, health and perceived service
quality). Third, we analyze the impact of these coping strategy clusters in relation
to the work conditions with regard to outcomes in health and perceived service
quality.
Coping strategies
Contemporary coping research related to job stress draws to a large extent on the
contextual and cognitive approach presented by Lazarus (1966). His approach set a
new course in including thoughts and behaviors that people use to manage
demands that are appraised as stressful, and also addressing the problems causing
the distress. Several researchers (Mickelson, Lyons, Sullivan, & Coyne, 2001; Peiro
´,
2008; Pienaar, 2008) have pointed out that the models used to understand relevant
phenomena at work, concerning coping, stress and employee health, need to be
enlarged and redefined. A main critique concerns the individualistic perspective
that has focused on the individual’s appraisal of stress, isolated from its social
context (Hobfoll, Dunahoo, Ben-Porath, & Monnier, 1994; La
¨nsisalm, Peiro
´,&
Kivima
¨ki, 2000; Newton, 1995; Mickelson et al., 2001). Accordingly, one of the
emerging research fields being emphasized presently, is the development and use of
more socially contextualized theoretical models and approaches that move beyond
Astvik and Melin 339
simplistic models not acknowledging the differences between the contexts in which
coping takes place.
Another issue that is emphasized in the ongoing coping debate is how to bridge
the gap between coping research and practice, how to find models and methodol-
ogies that allows us to make practical use of the extensive amount of coping
research that are carried out. Several researchers argue the need for using alterna-
tive approaches as a complement to traditional methods (Dewe & Trenberth, 2004;
Lazarus, 2000). According to Dewe and Trenberth (2004) coping research need to
make better use of transactional theories of stress and coping linking the individual
to the environment, and to adopt measurement methods that are more ecologically
sensitive, person- and meaning-centered and narrative in application.
We believe that a good understanding of the coping process and the factors
relevant to this process is crucial for interventions such as risk analysis and pre-
vention as well as the promotion of employee health and organizational perfor-
mance. With this follows a more contextualized approach, where the individual’s
appraisal of stress is linked to its social context. The present study employed a
socially contextualized approach in the sense that it was based on the stressors and
means of coping relevant to the study population, consisting of human service
workers within social work, child care and elderly care. The coping strategy mea-
surements in the questionnaire, that this article is based on, derive from results
from a comprehensive qualitative, narrative analysis of a sample of in-depth
interviews conducted within the project. The qualitative part of the project included
59 individual interviews and nine group interviews with another 51 individuals.
During the interviews, the respondents were asked to describe a number of
situations at work that they perceived as difficult and stressful and were also
given supplementary questions about how they dealt with each particular situation.
The vast majority of respondents described stressful situations concerning different
kinds of imbalance situations at work, such as periods of excessive workload and
insufficient time or uncertainty about how to handle difficult problems. The results
of the interviews were grouped, categorized and finally operationalized into a set of
items regarding the respondents’ ways of coping with excessive workload (a more
detailed description of the items is found under the headline ‘Method’).
Work conditions and coping
In the present study, we used the Job Demands-Resource (JD-R) model (Bakker &
Demerouti, 2007; Demerouti et al., 2001) to investigate demands and resources,
adding also the impact of different coping strategies (on outcomes in health and
service quality) adopted by individuals to manage the imbalance between resources
and work demands. The JD-R model incorporates specific (contextualized) risk
factors as well as addresses general work related problems associated with the
imbalance between work demands and resources. The JD-R model is consistent
with the demand-control model (Karasek & Theorell, 1990) in proposing that job
resources may buffer the impact of job demands on job strain (Bakker, Demerouti,
340 Journal of Social Work 13(4)
Taris, Schaufeli, & Schreurs, 2003), but expands this model by claiming that several
different job resources can act as a buffer for several different job demands. The
risk factors are classified into two general categories: job demands and job
resources. Job demands refer to those physical, psychological, social or organiza-
tional aspects of the job, requiring sustained physical and/or psychological effort or
skills and therefore being associated with certain physiological and/or psycholog-
ical costs. Job resources refer to those physical, psychological, social or organiza-
tional aspects of the job that are either functional in achieving work goals and
reducing job demands (as well as the associated physiological and psychological
costs), or stimulating personal growth, learning and development.
Based on the definition offered by Lazarus and Folkman (1984), coping, in this
study, is defined as constantly changing cognitive and behavioral efforts to manage
the internal and external demands of transactions that strain or exceed a person’s
resources. The aim of the present study was to identify different patterns of coping
strategies adopted by human service workers to manage the imbalance between
work demands and resources. This pattern approach is in line with Lazarus (1999),
stating that coping strategies should never be thought of in either/or terms, but as a
complex of interconnected thoughts and actions. The intention of the coping
strategy measure in the present study was to capture coping behavior in relation
to relevant and important stressors in the specific context of the studied groups and
at the same time add to a broader, generalized understanding of the dimensions of
coping, significant to human service work. Using the JD-R model allowed
incorporation of specific risk factors in the study population and at the same
time addressing general work related problems associated with the imbalance
between work demands and resources that is a wide spread problem in human
service work.
Method
Data collection and sample characteristics
The participants were employees in the public welfare sector taking part in a
larger project that included nine case studies concerning strategies for health
and service quality. Data for the present study were collected during the period
spring to autumn 2008 by means of a questionnaire to the total staff (n¼247)
at the nine workplaces; three workplaces within home care work, three within
social work and three within child care work. Participation was voluntary but
encouraged by the employers. Completed questionnaires were received from
195 (79%) employees after three reminders. Of these, 192 persons had com-
pleted all questions concerning strategies and this material was used for cluster
analysis of coping strategies. The majority of the participants were females
(91%) and the mean age was 46.7 years (SD ¼11.0). Home care workers
entailed 35.6 percent of the participants, 38.2 percent were social workers
and 26.2 percent were child care workers.
Astvik and Melin 341
Measures
The questionnaire was designed to capture demographic characteristics, such as
age, sex and information about employment characteristics (permanent/temporary,
type of work and occupational tenure, full-time/part-time employment), working
part-time because full-time is perceived as too demanding (yes/no), days on sick
leave (none or less than one week/one week or more), occasions of sick leave (none
or one to three times/four times or more), vacation instead of sick leave (never/one
time or more) and sickness presence (never or once/two times or more). The ques-
tionnaire also included measures of health behavior such as smoking (yes/no),
exercise (never or a few days per month/once a week or more) and physical activity
(never or a few days per month/once a week or more).
Finally, the questionnaire also measured experiences of the working conditions,
strategies to cope with excessive work demands and difficulties in work, perceived
service quality and health. A detailed account for these variables is presented
below.
Strategies. Coping strategies were assessed by 13 items regarding the respondents’
ways of handling excessive workload and difficulties at work. The items were con-
structed based on the results from an extensive qualitative analysis of 59 individual
in-depth interviews, relatively evenly distributed among the nine workplaces, and
nine group interviews with another 51 individuals. In the interviews the respon-
dents were asked to describe a number of situations in work that they perceived as
difficult and stressful, supplemented with questions about how they dealt with each
particular situation. As already mentioned, the vast majority of respondents
described stressful situations concerning different kinds of imbalance situations
in work, such as periods of excessive workload and insufficient time or uncertainty
about how to handle difficult problems. The content of the interviews were initially
grouped in different categories regarding strategies to cope with excessive work-
load, work conditions, health and service quality. After further groupings within
each category, several distinct themes emerged. Regarding coping strategies, the
themes that were identified included giving voice, seeking support, using compen-
satory efforts and trying in different ways to reduce work demands. Each of these
themes, in turn, comprised several sub-themes that in the next step were operatio-
nalized into a set of items regarding the respondents’ ways of handling excessive
workload. To increase validity, all groupings and categorizations were carried out
by two persons independent of each other. In total, 13 items were constructed for
the questionnaire. Examples of items were ‘In order to complete my job I skip
lunch and breaks’, ‘I voice my dissatisfaction about problems at work’, ‘I ask
colleges for help when workload is high’ and ‘I lower the demands for quality in
order to manage the work tasks to be done’. All strategies were assessed using
scales ranging from 1 (very seldom/never) to 5 (very often/always). The qualitative
analysis of the interviews revealed that respondents often tended to use several
types of strategies. Therefore, a cluster analysis was applied to identify patterns
342 Journal of Social Work 13(4)
of coping strategies among the respondents. For the predictive analysis of effects of
strategy profiles on health and service quality, two categorical dummy variables
were constructed following recommended procedures (Aiken & West, 1991).
Compensatory and quality reducing strategies (1 ¼compensatory and quality
reducing cluster, 0 ¼others) self-supporting strategies (1 ¼self-supporting cluster,
0¼others), with the voice and support seeking cluster as the comparison group.
Work conditions. A principal component analysis with varimax rotation of items
measuring work demands and resources indentified three factors: complex work
demands,social support and control of resources and demands. The items primarily
originate from the QPS Nordic questionnaire (Dallner et al., 2000) but also from
the Swedish Work Environment Survey (Arbetsmiljo
¨verket, 2008). All items were
assessed using scales ranging from 1 (very seldom/never) to 5 (very often/always).
A mean value index of six items was constructed to measure demands, covering
complex work demands, and two indexes to measure resources, consisting of a mean
value of three items for social support and a mean value of two items for control of
resources and demands. Originally, 17 items were included in the principal compo-
nent analysis, but six of the items were discarded from the final analysis because of
double loading or not loading highly on either factor. All items with factor loadings
are accounted for in Table 1.
Health indicators. The present study includes three different indicators of health.
First, a shortened version of the recuperation-scale was used (Aronsson,
Svensson, & Gustafsson, 2003; Gustafsson, Lindfors, Lundberg, & Aronsson,
2006). In the present study, the index measuring lack of recuperation and fatigue
was used, including six of the eight items that measure mental fatigue and recu-
peration only. The internal consistency (Cronbach’s a) of the original scale varied
from .85 (Gustafsson et al., 2006) to .84 (von Thiele, Lindfors, & Lundberg, 2006).
The Cronbach’s afor the shortened version of the recuperation scale in the present
study was .86. Second, a single item was used to measure work related sleeping
problems: ‘Have you during the last three months had difficulties sleeping due to
thoughts of work?’ Recuperation and work related sleeping problems were assessed
using scales ranging from 1 (very seldom/never) to 5 (very often/always). Third,
stress related symptoms were measured using an index (a¼.90) assessing a number
of psychological symptoms, that is, irritability and worry, symptoms most com-
monly reported by women and men in population-based studies and used in a
Swedish longitudinal cohort study (SLOSH), focusing on the relation between
work organization, work conditions and well being (Kinsten et al., 2006). The
scale measured symptoms during the last three months with responses ranging
from 1 (never) to 5 (every day).
Service quality. A mean value index of four items was constructed to measure sub-
jective service quality. The items were the following: ‘Do you think your work
group accomplishes a good quality for your clients?’ ‘Do you think that your
Astvik and Melin 343
clients are satisfied with the quality they receive?’ ‘Do you feel proud of the work
that you and your colleges accomplish?’ ‘Are you satisfied with the quality of your
own work?’ Cronbach’s afor the index was .83. The responses ranged from 1 (very
seldom/never) to 5 (very often/always).
Descriptive statistics (means, standard deviations and inter-correlations) for the
variables are presented in Table 2.
Statistical analyses
The first objective of this study, to identify different patterns of coping strategies,
was addressed using cluster analysis. The Quick cluster procedure in SPSS (Version
16.0) was used to perform a cluster analysis of the 13 items in the coping strategy
measure. A K-means cluster analysis identifies homogeneous groups of cases based
on selected characteristics; in this case the strategy items, using an algorithm based
Table 1. Principal component analysis of questionnaire items concerning work conditions.
Items and factor loadings
Component
Item no. 1 2 3
Complex work demands
13c Psychologically straining work .569 .391 .298
13e Work demands to learn new things .605 .220 .150
13f Perform work that require more education .613 .183
13g Have to re-organize and change priorities .703 .430
13h Face complex tasks and problems in work .727 .107 .409
13i Work include many different tasks .739
Social support
13n Support from colleges .835
13o Support from supervisors .739 .126
13q Can openly discuss difficulties .725 .257
Control of resources and demands
13j Influence over resources .123 .149 .860
13k Influence over demands .136 .243 .811
Eigenvalue 2.68 2.10 1.95
Percentage of variance explained 24.33 19.05 17.72
Cronbach’s alpha .79 .81 .77
Notes: Loadings in bold indicate the variables contributing most to a component, and were also the ones used
for index construction.
Loadings <.10 are not shown in the table. Response scales ranged from: (1) All most never/never to (5) Very
often/always.
344 Journal of Social Work 13(4)
Table 2. Correlations, means and standard deviations for all variables, and afor scales (N¼191)
123456789101112
1. Age 1.00
2. Occupational tenure .60** 1.00
3. Complex work demands .11 .17* 1.00
4. Control of resources .13 .11 .42** 1.00
5. Support .16* .24** .19* .36** 1.00
6. Compensatory and quality
reducing strategies
.01 .18* .51** .34** .22** 1.00
7. Self-supporting strategies .07 08 38** .14 .06 .40** 1.00
8. Voice and support seeking
strategies
.06 .23** .17* .21** .26** .63** 46** 1.00
9. Tiredness/lack of recuperation .23** .30** .43** .29** .40** .41** 19* .24** 1.00
10. Stress related symptoms .19* .24** .34** .27** .37** .34** .07 .27** .65** 1.00
11. Sleeping difficulties due
to work
.02 .14* .43** .25** .31** .43** .13 .31** .56** .62** 1.00
12. Subjective quality in work .10 .23** .49** .45** .47** .47** .24** .25** .51** .36** .39** 1.00
Mean 46.66 14.76 3.35 2.36 3.92 .36 .22 .42 2.41 1.99 2.07 4.05
Standard deviation 11.02 10.59 .65 .97 .75 .48 .42 .49 .69 .78 1.18 .61
Reliability (a) – – .79 .77 .71 – – – .86 .90 – .83
**p.01; *p.05.
Note: Scale range 0–1 (variables 6, 7 and 8 [for these variables the mean value symbolizes the proportion scoring 1]) 1–5 (variables 3–5. 9–12).
Astvik and Melin 345
on nearest centroid sorting. This means that a case is assigned to the cluster with
the smallest distance between the cluster centroid and the case (Aldenderfer &
Blashfield, 1991). K-means cluster analyses require specification of the number of
clusters. Based on results from the qualitative analysis of the interviews, a three-
cluster solution was assumed. Additionally, cluster solutions with two, four and
five clusters were calculated. However, the three cluster solution generated the most
homogenous profiles. Furthermore, a residual analysis (Stem-and-Leaf and
Boxplot) was used to detect outliers, resulting in one extreme case (5.0) being
omitted from further analysis.
In order to compare differences between the three strategy profiles a Univariate
analysis of variance (ANOVA) was performed on background characteristics,
work conditions, health indicators and perceived service quality. F-tests and
Sheffe
´post hoc tests were furthermore performed. For variables assessed as cate-
gorical variables, a
2
test was used.
For the predictive analysis of effects of working conditions and strategy profiles
on health and perceived service quality, a hierarchical regression analysis was per-
formed on each of the dependent variables. The data were screened for outliers,
multicollinearity and tolerance levels to ensure that the requirements for multiple
regressions were met.
In each of the performed regressions, individual background variables (age and
occupational tenure) were first entered into the model to control for their effect on
the dependent variables. In the second step, complex work demands, social support
and control of resources and demands were included in the model in order to inves-
tigate whether differences in these working conditions explained variances in health
and service quality while controlling for age and occupational tenure. In step three,
strategies were added to the model in order to test whether strategies had any
additional significant effect on the outcome variables.
Results
Strategy profiles
The cluster analysis resulted in three clusters. Figure 1 describes the patterns of the
three cluster profiles in terms of their mean values of the items in the strategy
measure. Mean values for all items included in the strategy measure differed sig-
nificantly between the three clusters. The three strategy profiles were: compensatory
and quality reducing,self-supporting and voice and support seeking. The compensa-
tory and quality reducing cluster consisted of 68 persons and was characterized by
an extensive use of compensatory strategies. They included: working more inten-
sively, skipping lunches and other breaks, bringing work home on leisure time and
taking on more work than they had time for. Individuals in this cluster also tended
to think about work during their leisure time. All these compensatory strategies
were significantly more common in this cluster profile compared to the other two.
In addition to these typically compensatory strategies, they also lowered their
346 Journal of Social Work 13(4)
demands for quality and asked supervisors to prioritize among the work tasks
significantly more often than the two other profile groups.
The voice and support seeking cluster consisted of 80 persons and was – rather
than using compensatory strategies – characterized by giving voice and more
proactively asking others for help. These strategies involved seeking support
from colleagues when facing difficulties at work and also asking colleagues for
help when workload was high. Individuals in the voice and support seeking cluster
were also seeking help from supervisors to prioritize among work tasks, objecting
when the workload was too excessive and giving voice to dissatisfaction as well as
actively trying to find solutions to problems in the work group.
The self-supporting cluster consisted of 43 persons and was characterized by
using fewer coping strategies than the other two cluster profiles. The individuals in
this cluster did not use compensatory strategies very often, except for working
more intensively and thinking about work during leisure time (significantly more
often than the individuals in the voice and support seeking cluster but significantly
less often than the individuals in the compensatory and quality reducing cluster).
Moreover, they did not prioritize among work tasks or reduced the quality. The
label self-supporting refers to this strategy profile’s tendency not to ask colleagues
or supervisors for support, help or to prioritize. In addition, they did not voice
dissatisfaction or object to excessive work demands.
Comparing demographic factors, there was no significant difference between
cluster groups regarding age (F(2, 188) ¼.557, p¼.574), gender (
2
(2) ¼1.390,
p¼.499) or type of employment (
2
(2) ¼1.870, p¼.393) (permanent/temporary).
Almost half of the self-supporting cluster comprised part-time workers (48.8%),
compared with one third of the voice and support seeking cluster (31.2%) and only
1.00
1.50
2.00
2.50
3.00
3.50
4.00
4.50
5.00
Compensatory and quality reducing n=68 Self-supporting n=43 Voice and support seeking n=80
Figure 1. Mean value of strategy items for three clusters with distinct stragegy profiles.
Astvik and Melin 347
one fifth of the compensatory and quality reducing cluster (22.1%) (
2
(2) ¼8.720,
p¼.013). Among those working part-time, around 50–60 percent of the three clus-
ters did so because full-time was perceived as too demanding (
2
(2) ¼.480,
p¼.787).
The individuals in the compensatory and quality reducing cluster had signifi-
cantly fewer years of experience in the occupation than those in the voice and
support seeking cluster (compensatory and quality reducing cluster mean ¼12.4
years, SD ¼9.1 years; voice and support seeking cluster mean ¼17.9 years,
SD ¼10.8 years; self-supporting cluster mean ¼13.4 years, SD ¼11.5 years) (F(2,
189) ¼5.565, p¼.004]. A greater proportion of the compensatory and quality
reducing cluster members was employed in the social services (63.2%), whereas
the self-supporting cluster included a greater proportion of home care workers
(60.5%) and the voice and support seeking cluster included a higher proportion
of child care workers (41.2%) (
2
(4) ¼40,328, p¼.000).
Regarding the number of days (
2
(2) ¼.473, p¼.789) and occasions (
2
(2) ¼4.114, p¼.128) on sick leave, there were no significant differences between
the cluster groups. However, a greater proportion of the individuals in the com-
pensatory and quality reducing cluster tended to work despite sickness (64.7%)
compared to the voice and support seeking cluster (32.9%) and self-supporting
cluster (46.5%) [
2
(2) ¼14.834, p¼.001]. The compensatory and quality reducing
cluster further differed from the other two cluster groups in taking vacation instead
of sick leave; 41.2 percent of these individuals had taken vacation instead of sick
leave more than once, compared to 18.2 percent of the self-supporting cluster and
20.3 percent of the voice and support seeking cluster (
2
(2) ¼10.196, p¼.006).
Concerning health behaviors, there were no significant differences between the
cluster groups regarding physical activity (
2
(2) ¼1.880, p¼.391], exercise (
2
(2) ¼.313, p¼.855) or smoking (
2
(2) ¼3.610, p¼.164). More than half of the
full sample (51.4%) exercised a couple of times a week, 76.8 percent were doing
some kind of physical activity a couple of times a week and three-quarters (75.3%)
of the full sample were non-smokers.
Table 3 presents mean levels for background characteristics, work conditions,
health and perceived service quality of the three strategy profiles. It also reports the
results of the tests regarding differences between the compensatory and quality
reducing, the self-supporting and the voice and support seeking cluster. The
F-tests revealed significant differences for all job conditions, while the Sheffe
´post
hoc tests showed that the level of complex work demands differed between all three
clusters with the highest level in the compensatory and quality reducing cluster and
the lowest in the self-supporting cluster. The post hoc tests also revealed that the
compensatory and quality reducing cluster had lower levels regarding control of
recourses and demands than both the self-supporting and voice and support seek-
ing cluster. The compensatory and quality reducing cluster also expressed lower
levels of social support than the voice and support seeking cluster.
Furthermore, the F-tests displayed significant differences regarding health, that
is, lack of recuperation/tiredness, stress related symptoms and sleeping difficulties
348 Journal of Social Work 13(4)
due to work. The Sheffe
´post hoc tests uncovered that individuals in the compen-
satory and quality reducing cluster had sleeping problems, stress related symptoms
as well as experienced tiredness/lack of recuperation significantly more often than
individuals belonging to both the self-supporting and voice and support seeking
cluster. Finally, the F-tests also disclosed significant overall differences regarding
service quality. The post hoc tests unveiled that individuals belonging to the com-
pensatory and quality reducing cluster were less satisfied with quality than individ-
uals in both the self-supporting and the voice and support seeking cluster.
Predicting health and service quality
Table 4 presents the results of hierarchical regression analyses predicting health
and service quality. In step one, the background variables displayed a significant
association with lack of recuperation (R
2
¼.07, p<.001) but none of the back-
ground characteristics reached significance. The second step, in which the working
Table 3. Test for mean differences in background characteristics, working conditions, health
and perceived service quality
Compensatory
and quality
reducing (1)
Self-
supporting
(2)
Voice and
support
seeking
(3) Total
Univariate
F
Group
comparisons
N68 43 80 191
Background
characteristics
Age 46.83 45.23 47.58 46.79 0.56
Occupational tenure 12.35 13.36 17.93 14.92 5.57** 1-3
Working conditions
Complex work
demands
3.77 2.89 3.19 3.34 37.50*** 1–2; 1–3;
2–3
Control of resources
and demands
1.92 2.61 2.59 2.36 11.65*** 1–2; 1–3
Support 3.70 3.83 4.16 3.92 7.57*** 1–3
Health indicators
Tiredness/lack of
recuperation
2.79 2.17 2.20 2.40 19.34*** 1–2; 1–3
Stress related
symptoms
2.34 1.88 1.73 1.98 12.55*** 1–2; 1–3
Sleeping difficulties
due to work
2.75 1.79 1.63 2.06 21.81*** 1–2; 1–3
Quality 3.66 4.33 4.23 4.06 26.48*** 1–2; 1–3
***p.001; ** p.01.
Astvik and Melin 349
conditions were added, demonstrated a rather large increase in explained variance
(R
2
change ¼.22, p<.001). Complex work demands were positively related and
social support were negatively associated with lack of recuperation. In step three,
the coping strategies were added (with the voice and support seeking cluster as the
comparison group), significantly increasing the overall explained variance of the
model (R
2
change ¼.04, p<.01). Those in the compensatory and quality reducing
cluster, experienced lack of recuperation significantly more often, compared to
those in the voice and support seeking and the self-supporting cluster. Regarding
stress related symptoms, background characteristics in step one exhibited a signif-
icant association (R
2
¼.06, p<.01), but none of the separate background charac-
teristics reached significance. With the background characteristics being held
constant, there was a significant increase in the explained variance when job con-
ditions were added in step two (R
2
change ¼.15, p<.001). Social support was
negatively correlated to stress related symptoms. In step three, strategies signifi-
cantly increased the explained variance (R
2
change ¼.02, p<.05). Again, the com-
pensatory and quality reducing cluster reported stress related symptoms
significantly more often than the voice and support seeking cluster as well as the
self-supporting cluster.
Table 4. Predicting health and subjective quality in work. Standardized regression coefficients
(N¼191)
Lack of
recuperation
last step
Stress related
symptoms b
last step
Sleeping
difficulties
blast step
Quality
last step
Step 1 Background characteristics
Age .12 .12 .06 .05
Occupational tenure .08 .07 .02 .10
Adjusted R
2
change .07*** .06** .01 .04**
Step 2 Working conditions
Complex work demands .27*** .17 .29*** .19**
Control of resources and demands .03 .01 .03 .17**
Social support .26*** .26*** .19** .30***
Adjusted R
2
change .22*** .15*** .19*** .33***
Step 3 Coping strategies
Compensatory and quality reducing .21** .20* .28*** .20**
Self-supporting .05 .05 .09 .10
Adjusted R
2
change .04** 02* .04** .04***
Adjusted total R
2
.33 .23 .24 .41
***p.001; **p.01; *p.05.
350 Journal of Social Work 13(4)
Sleeping difficulties were not significantly correlated to background variables.
However, in step two, when working conditions were added as well, the explained
variance reached significance (R
2
change ¼.19, p<.001). Complex work demands
were positively related to sleeping difficulties, while social support was negatively
associated with sleeping difficulties. In step three, the coping strategies were added,
thereby increasing the overall explained variance of the model (R
2
change ¼.04,
p<.01). Individuals belonging to the compensatory and quality reducing cluster
reported sleeping difficulties significantly more often than those in the other two
coping clusters.
In step one, the background variables are significantly associated with perceived
service quality (R
2
¼.04, p<.01), but as in the predictions of health, none of the
separate background characteristics reached a significant correlation. The second
step, in which the working conditions were added, exhibited a large increase in
explained variance (R
2
change ¼.33, p<.001). Complex work demands were neg-
atively related, while social support and control of recourses and demands were
positively associated with service quality. In step three, the coping strategies were
added, which significantly increased the overall explained variance of the model (R
2
change ¼.04, p<.001). Those in the compensatory and quality reducing cluster
reported significantly less satisfaction with service quality compared to those in the
voice and support seeking and the self-supporting clusters.
Discussion
The intention with the coping strategy measures was to investigate coping behavior
in relation to relevant and important stressors in the specific context of the studied
groups. The approach adopted in this study, that is, identifying and comparing
homogenous patterns of coping strategies rather than studying coping as single
variables or factors, was based on previous research stating that coping strategies
never should be thought of in either/or terms, but as a complex of interconnected
thoughts and actions (Lazarus, 1999). The use of cluster analysis was furthermore
justified by the results of the qualitative analysis of the interviews, revealing that
respondents often tended to use several types of strategies. Analysis of the linear
association between single variables, such as comparisons between emotion-
focused versus problem-focused coping and different outcome variables has been
widely endorsed in coping research. However, questions still remain about possible
consequences of different patterns of strategies on the outcome variables. There is
also a psychometric issue related to this, which concerns the presumption that the
factors in a (multifactorial) coping scale should be independent of one another.
Folkman and Moskowitz (2004) point out that distinct kinds of coping (such as,
for example, problem-focused versus meaning-focused coping) conceptually seem
to be entwined, suggesting that these two forms of coping facilitate each other.
Thus, using for example factor analysis and expecting that the coping factors
should be uncorrelated may actually result in a reduction of the validity of the
measure and also mask important differences within the categories. As described in
Astvik and Melin 351
the introduction, we are not using an established coping inventory or categoriza-
tion for our coping strategy measure. In order to uncover coping strategies relevant
to the studied population, the most appropriate solution rather seemed to be to
start with the narrative approach, and from that point define items relevant for the
quantitative measure.
In our study, we identify three coping strategy patterns connected to excessive
work demands and difficulties at work, namely the compensatory and quality reduc-
ing cluster, the self-supporting cluster and the voice and support seeking cluster. The
compensatory and quality reducing cluster was characterized by an extensive use of
compensatory strategies, but also lowering the demands of quality. In contrast, the
voice and support seeking cluster was characterized by giving voice and asking
colleagues and supervisors for help and support. Instead of using compensatory
strategies, they objected when the workload was to excessive, sought help to pri-
oritize work tasks and actively tried to find solutions to problems in the work
group. Finally, the self-supporting cluster was characterized by using few coping
strategies and appeared self-supporting in their tendency not to ask colleagues or
supervisors for support, help or to prioritize. Furthermore, they did not voice
dissatisfaction or object to excessive work demands. At the same time, the use of
compensatory strategies was limited to working more intensively and thinking
about work during leisure time. Noticeable however, is that almost 50 percent of
the persons in the self-supporting cluster worked part-time compared to a third in
the voice and support seeking cluster and only one fifth in the compensatory and
quality reducing cluster. As around 50–60 percent of those working part time did so
because they perceived full-time work too demanding, one must also consider that
the self-supporting cluster included individuals that were using part-time as a strat-
egy to reduce their total workload, which may be an alternative for using strategies
at work.
There are also interesting differences between the clusters regarding how they
handled sickness and absence from work. The compensatory and quality reducing
cluster distinguished from the other two in a negative way, by tending to work
despite being sick and taking vacation instead of sick leave. Sickness presenteeism
can also be conceived as a type of compensatory strategy. Previous studies have
reported that working conditions increasing the probability of sickness presentee-
ism are: time pressure, insufficient resources to accomplish a good quality as well as
the work not being done by someone else while the employee is on sick leave
(Aronsson & Gustafsson, 2005; Aronsson, Gustafsson, & Dallner, 2002). Recent
research shows that sickness presenteeism appears to be an independent risk factor
in terms of future poor general health (Bergstro
¨m et al., 2009a) as well as future
sick leave (Bergstro
¨m, Bodin, Hagberg, Aronsson, & Josephson, 2009b). Despite
these behavioral differences, there were no differences between the clusters in health
behaviors such as physical activity, exercise or smoking.
The results of the comparative analyses indicates that the most interesting
clusters from a health and quality perspective, is on the one hand the com-
pensatory and quality reducing cluster as being a ‘risk group’, and on the
352 Journal of Social Work 13(4)
other hand the voice and support seeking cluster that seems to have adopted
more beneficial strategies. Overall, the compensatory and quality reducing clus-
ter differed from the other two clusters, but also the voice and support seeking
cluster singled out from one or both of the other cluster profiles in some
aspects. One interesting aspect concerns number of years in the occupation,
where the compensatory and quality reducing cluster had significantly fewer
years in the occupation than the voice and support seeking cluster. With
extended tenure, the possibility of establishing social relationships and devel-
oping more extensive networks also increase. In the present study, fewer years
in the occupation of the compensatory and quality reducing cluster, may
explain their overall more disadvantageous strategies.
Concerning the work conditions, the highest level of complex work demands
was found in the compensatory and quality reducing cluster. In addition, the
compensatory and quality reducing cluster had less control of resources and
demands than the other two strategy clusters, and also reported less social
support than the voice and support seeking cluster. Drawing on the JD-R
model, an imbalance between job demands and resources may result in nega-
tive outcomes in health. As presented above, post hoc tests disclosed that the
individuals in the compensatory and quality reducing cluster suffered from
sleeping problems, stress related symptoms and experienced tiredness/lack of
recuperation more often than the individuals in the other two coping clusters.
According to the JD-R model, several different job resources may play the role
of buffer to several different job demands. However, social support is probably
the most well known situational variable found to have both a direct and
buffering effect on job strain (Viswesvaran, Sanches, & Fisher, 1999). In our
study, the voice and support seeking cluster reported the highest levels of
social support.
The health consequences associated with compensatory strategies may also be
analyzed within the cognitive-energetical framework of Hockey’s (1997) compen-
satory regulatory-control model. According to this model, employees under high
workload face a compromise between the protection of their performance goals
and the mental effort that has to be invested in order to accomplish these goals.
When job demands increase, compensatory effort has to be mobilized to deal with
the increased demands whilst maintaining performance levels. This additional com-
pensatory effort is associated with physiological and psychological costs such as
increased sympathetic activity. Continuous mobilization of compensatory effort
drains the employee’s energy and might therefore lead to burnout and ill-health
(Hockey, 1997). An alternative response to excessive work demands according to
this model is to adopt strategies involving downwards adjustment of performance
goals. Hockey’s model states that the regulation process of excessive work demands
is intra-individual and does not include the option of for example using external
resources such as help from colleagues in handling the workload, which in our
study occurred within the voice and support seeking cluster. Nevertheless, in accor-
dance with Hockey’s model, the individuals in the compensatory and quality
Astvik and Melin 353
reducing cluster obviously used compensatory efforts in order to achieve their work
goals and consequently suffered from reduced health. But the response according
to Hockey’s model, adopting strategies involving downwards adjustment of per-
formance goals, was in our study used only in combination with the compensatory
effort, not as an alternative. However, according to Hockey (1997) fatigue is a
natural consequence of compensatory effort, but when it is associated with chronic
stress fatigue may also reflect a baseline reduction in adaptive capacity. The sen-
sitivity of fatigue to high work demands and most kinds of illness might according
to Hockey indicate an adaptive role in shifting behavior towards less effortful
strategies. The performance–cost trade-off depicted in the model, that is, the rela-
tionship between compensatory performance protection and increased sympathetic
activity has been found in laboratory studies (Lundberg & Frankenhauser, 1978) as
well as field studies (Gustafsson, 2008; Rissler & Elgerot, 1978; Rissler & Jacobson,
1987). Thus, based on previous research and theory, the relationship between
coping patterns and various health indicators were in an expected direction, and
strengthen the validity of the study.
Strategies in an organizational context
Comparisons between the identified coping clusters displayed differences in
coping patterns between the three studied organizations. A greater proportion
of the compensatory and quality reducing cluster were social workers, whereas
the self-supporting cluster included a greater proportion of home care workers
and the voice and support seeking cluster included a higher proportion of child
care workers. Looking at these coping strategies in an organizational context,
the differences appear coherent. For instance, the child care work in this study
was organized in teams of three to four employees, attending to and having a
mutual responsibility for a group of children. Communication and cooperation
between team members were of vital importance to accomplish the work and
to achieve quality. Home care work on the other hand, was to a very large
extent based on individual performances. After a short meeting in the morning
where tasks and clients were distributed in the group, the rest of the work day
was spent without colleagues in the homes of the clients. The home care work
consisted of a relatively small amount of time interacting with colleagues and
supervisors in group meetings and similar activities. Thus, the possibilities for
voice and support seeking strategies were limited. Finally, the social workers
were organized in specialized teams and also had a number of different meet-
ings with colleagues such as team meetings, team supervision, workplace meet-
ings, etc. This kind of organization ought to give room for voice and social
support strategies, but this was probably hampered by the extreme workload
that characterized the work situation. When all team members are equally
overloaded with work, the actual scope for voice and support seeking strategies
may be perceived as minimal and the use of compensatory and quality reduc-
ing strategies therefore considered as the only alternative.
354 Journal of Social Work 13(4)
A greater proportion of the compensatory and quality reducing cluster members
in this study were social workers and it was also in this cluster that the highest
levels of complex work demands were found. These results are coherent with other
empirical studies. Social work stands out as a particular exposed occupational
group among human service professions (Tham & Meagher, 2008) and several
studies of work conditions in social work have found that this group experience
high workload (Balloch et al., 1998; Coffey et al., 2004, 2009; Collings & Murray,
1996; Huxley et al., 2005). In Sweden, social workers are one of the occupations
with the largest proportion of stress-related disorders (National Work
Environment Authority & Statistics Sweden, 2010).
Predictions of health and quality
The third objective of the present study was to investigate whether the use of
different strategies contributed to the prediction of employee health, in terms of
stress related symptoms, tiredness/lack of recuperation and sleeping problems due
to work as well as employees perception of service quality. The results confirm that
strategies affected all health outcomes as well as perceived service quality. The use
of compensatory and quality reducing strategies was without exception negatively
related to health, even though known explanatory factors such as work demands
and recourses as well as background characteristics were taken in to account.
Limits of the study may be the cross-sectional design that possibly restricts
causal interpretation as well as the risk of common method variance (Podsakoff,
McKenzie, Lee, & Podsakoff, 2003). Self-reports collected by means of question-
naires are considered as appropriate measures to reflect individual perception of
work conditions, coping strategies, well-being and perceived quality, but they carry
the risk of common method variance, artificially inflating the association between
the measured constructs. In order to reduce potential risks of common method
variance, we followed recommendations for suitable questionnaire techniques such
as anonymity, instructing participants that there is no right or wrong answers as
well as changing the response format (Podsakoff et al., 2003; Spector, 2006).
Nevertheless, future studies may benefit from employing different design strategies
in order to further investigate the relationship of the variables presented here
(Spector, 2006). The cross-sectional design in this study does not allow for any
causal conclusions. However, the interpretation of a causal relationship between
compensatory strategies and reduced health is theoretically and empirically sup-
ported by similar studies (Gustafsson, 2008; Hockey, 1997). For future research, a
longitudinal design would indeed contribute to the understanding of how strategies
contextually develop in a long-term perspective. It could, for example, be hypoth-
esized, given no major contextual changes, that individuals maintain the more
beneficial voice and support seeking strategies while individuals with compensatory
and quality reducing strategies would change, since these strategies are not sus-
tainable in the long run. As high turnover and retention of staff in social work is a
significant international problem (Audit Commission, 2002; Glisson, Dukes, &
Astvik and Melin 355
Green 2006; Healy et al., 2009; Lidell, Donnegan, Goddard, & Tucci, 2006; Tham,
2007), an interesting question concerns whether the direction of change for those
with compensatory and quality reducing strategies would be either towards more
beneficial strategies or towards exit strategies, such as turnover or long-term sick
leave.
Conclusions and practical implications
Summing up, the contextual approach applied in this study adds to coping theory
by the identified compensatory and quality reducing strategies that are not
described previously in the coping literature. In this article we have presented
results that indicate that in order to promote good health and service quality, it
is vital to consider the decentralized and deregulated work conditions that, in sit-
uations with an imbalance between demands and resources, enforce the individual
to take on the responsibility for maintaining performance levels as well as being
compelled to compromise the standards of quality. The results illustrate the risk
that individuals start using compensatory efforts to deal with increased demands in
trying to maintain performance levels, and when the compensatory efforts are not
enough, the individual may also be compelled to make downward adjustments of
the aspired level of service quality. Applied in practice, the identified strategy clus-
ters might help detect and distinguish ‘risk behaviors’ from more beneficial strat-
egies. If, for example, compensatory and quality reducing strategies are frequent in
an organization, it might be an indicator that employees are pushing themselves
too hard in order to fulfill perceived obligations and duties. In this way, interven-
tions could be targeted to those in most need. The results also point toward the
importance of providing work conditions and organizational structures that allow
employees to voice their opinions and criticisms, as well as to collaborate and
support each other. For future research, there is a need for more research dealing
with interventions to find ways to create sustainable work environments that sup-
port employees in developing strategies that neither jeopardize their health nor
impair service quality. There is also a challenge to expand the knowledge about
the interplay between strategies and work environment factors that may restrict the
individual’s scope of action.
Ethical approval
According to the statues of the Central Ethical Review Board in Sweden between 1
January 2004 and 31 May 2008 (when the application of this study was submitted),
ethical approval was not necessary if the subject of the research was given infor-
mation about the research and consented to the research that concerned him or her.
Funding
The study was funded by a research grant from AFA Fo
¨rsa
¨kring [Insurance] (Dnr 060004).
356 Journal of Social Work 13(4)
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions.
Thousand Oaks, CA: SAGE.
Aldenderfer, M. S., & Blashfield, R. K. (1991). Cluster analysis. Thousand Oaks, CA:
SAGE.
Allvin, M. (2008). New rules of work: Exploring the boundaryless job. In K. Na
¨swall,
J. Hellgren, & M. Sverke (Eds), The individual in the changing working life (pp. 19–45).
New York: Cambridge University Press.
Arbetsmiljo
¨verket. (2008). Arbetsmiljo
¨n 2007 [The Work Environment 2007].
Arbetsmiljo
¨statistisk rapport 2008:4. Solna: The Swedish Work Environment Authority.
Aronsson, G., & Gustafsson, K. (2005). Sickness presenteeism: Prevalence, attendance-
pressure factors and an outline of a model for research. Journal of Occupational and
Environmental Medicine,47(9), 958–966.
Aronsson, G., Gustafsson, K., & Dallner, M. (2002). Sick but yet at work. An empirical
study of sickness presenteeism. Journal of Epidemiology & Community Health,54(7),
502–509.
Aronsson, G., Svensson, L., & Gustafsson, K. (2003). Unwinding, recuperation and health
among compulsory-school and high-school school teachers in Sweden. International
Journal of Stress Management,10(3), 217–234.
Audit Commission. (2002). Recruitment and retention: A public service workforce for the
twenty-first century. London: Audit Commission.
Bakker, A. B., & Demerouti, E. (2007). The job demands-Resources model: State of the art.
Journal of Managerial Psychology,22(3), 309–328.
Bakker, A. B., Demerouti, E., Taris, T., Schaufeli, W. B., & Schreurs, P. (2003). A multi-
group analysis of the Job Demands-Resources model in four home care organizations.
International Journal of Stress Management,10(1), 16–38.
Balloch, S., Pahl, J., & McLean, J. (1998). Working in the social services: Job satisfaction,
stress and violence. British Journal of Social Work,28(3), 329–350.
Bergstro
¨m, G., Bodin, L., Hagberg, J., Lindh, T., Aronsson, G., & Josephson, M.
(2009a). Does sickness presenteeism have an impact on future general health?
International Archives of Occupational and Environmental Health, DOI 10.1007/s00420-
009-0433-6.
Bergstro
¨m, G., Bodin, L., Hagberg, J., Aronsson, G., & Josephson, M. (2009b). Sickness
presenteeism today, sickness absenteeism tomorrow? A prospective study on sickness
presenteeism and future sickness absenteeism. Journal of Occupational and
Environmental Medicine,51(6), 629–638.
Coffey, M., Dugdill, L., & Tattersall, A. (2004). Stress in social services: Mental wellbeing,
constraints and job satisfaction. British Journal of Social Work,34(5), 735–746.
Coffey, M., Dugdill, L., & Tattersall, A. (2009). Working in the public sector. A case study
of social services. Journal of Social Work,9(4), 420–442.
Collings, J., & Murray, P. (1996). Predictors of stress among social workers: An empirical
study. British Journal of Social Work,26(3), 375–388.
Dallner, M., Lindstro
¨m, K., Elo, A.-L., Skogstad, A., Gamberale, F., Hottinen, V.,
Knardahl, S., & Ørhede, E. (2000). Anva
¨ndarmanual fo
¨r QPS Nordic [User’s guide for
the QPSNordic]. Arbetslivsrapport, 19. Solna: Arbetslivsinstitutet.
Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-
resources model of burnout. Journal of Applied Psychology,86(3), 499–512.
Astvik and Melin 357
Dewe, P., & Trenberth, L. (2004). Work stress and coping: Drawing together research
and practice. British Journal of Guidance & Counselling,32(2), 143–156.
European Agency for Safety and Health at Work. (2009). OSH in figures: Stress at work –
facts and figures. European risk observatory report 9. Luxembourg: Office for Official
Publications of the European Communities.
Folkman, S., & Moskowitz, J. T. (2004). Coping: Pitfalls and promise. Annual Review of
Psychology,55(1), 745–774.
Glisson, C., Dukes, D., & Green, P. (2006). The effects of the ARC organizational inter-
vention on caseworker turnover, climate, and culture in children’s service systems. Child
Abuse and Neglect,30(8), 855–880.
Gustafsson, K. (2008). Working conditions, compensatory strategies, and recovery.
Doctoral dissertation, Department of Psychology, Stockholm University.
Gustafsson, K., Lindfors, P., Lundberg, U., & Aronsson, G. (2006). Validering av fra
˚gor
avseende nedvarvning och a
˚terha
¨mtning – samband mellan salivkortisol och sja
¨lvskatt-
ningar [Validation of questions on recovery – relationships between salivary cortisol and
self-ratings of recovery]. Arbete & Ha
¨lsa [Work & Health].
Healy, K., Meagher, G., & Cullin, J. (2009). Retaining novices to become expert
practitioners: Creating career pathways in direct practice. British Journal of Social
Work,39(2), 299–317.
Hellgren, J., Na
¨swall, K., Sverke, M., & So
¨derfeldt, M. (Eds) (2003). New organizational
challenges for human service work. Organizational Psychology and Health Care, 3,
Munich: Rainer Hampp Verlag.
Hellgren, J., Sverke, M., & Na
¨swall, K. (2008). Changing work roles: New demands and
challenges. In K. Na
¨swall, J. Hellgren, & M. Sverke (Eds), The individual in the
changing working life (pp. 46–66). New York: Cambridge University Press.
Hobfoll, S. E., Dunahoo, C., Ben-Porath, Y., & Monnier, J. (1994). Gender and
coping: The dual axis model of coping. American Journal of Community Psychology,
22(1), 49–82.
Hockey, G. R. J. (1997). Compensatory control in the regulation of human performance
under stress and high workload: A cognitive-energetical framework. Biological
Psychology,45(1–3), 73–93.
Huxley, P., Evans, S., Gately, C., Webber, M., Mears, A., Pajak, S., Kendall, T., & Katona,
C. (2005). Stress and pressures in social work: The worker speaks. British Journal of
Social Work,35(7), 1063–1079.
Ha
¨renstam, A., Bejerot, E., Leijon, O., Sche
´ele, K., & Waldenstro
¨m, K. (2004).
Multilevel analyses of organizational change and working conditions in public and private
sector. European Journal of Work and Organizational Psychology,13(3), 305–343.
Ha
¨renstam, A., MOA Research Group (2005). Different development trends in working life
and increasing occupational stress require new work environment strategies. Work,24(3),
261–277.
Karasek, R., & Theorell, T. (1990). Healthy work. New York: Basic Books.
Kinsten, A., Magnusson Hanson, L., Hyde, M., Oxenstierna, G., Westerlund, H., &
Theorell, T. (2006). SLOSH. En riksrepresentativ studie av arbetsmiljo
¨och ha
¨lsa
[SLOSH. A nationally representative study of work environment and health]. Institutet
fo
¨r Psykosocial Medicin.
Lazarus, R. S. (1966). Psychological stress and the coping process. New York: McGraw-Hill.
Lazarus, R. S. (1999). Stress and emotion: A new synthesis. London: Free Association Books.
358 Journal of Social Work 13(4)
Lazarus, R. S. (2000). Towards better research on stress and coping. American Psychologist,
55(6), 655–673.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer
Publishing Company.
Lidell, M., Donnegan, T., Goddard, C., & Tucci, J. (2006). The state of child protection:
Australian child welfare and child protection development. Melbourne: National Centre for
the Prevention of Child Abuse.
Lundberg, U., & Frankenhauser, M. (1978). Psychophysiological reactions to noise as
modified by personal control over noise intensity. Biological Psychology,6(1), 51–59.
La
¨nsisalmi, H., Peiro
´, J. M., & Kivima
¨ki, M. (2000). Collective stress and coping in the
context of organizational culture. European Journal of Work and Organizational
Psychology,9(4), 527–559.
Mickelson, K., Lyons, R. F., Sullivan, M. J. L., & Coyne, J. C. (2001). Yours, mine and
ours: The relational context of communal coping. In B. R. Sarason, & S. Duck (Eds),
Personal relationships: Implications for clinical and community psychology (pp. 181–200).
Chichester: John Wiley.
National Work Environment Authority & Statistics Sweden. (2010). Work-related
disorders 2010. Arbetsmiljo
¨statistik, Rapport 2010:4. National Work Environment
Authority.
Newton, T. (1995). Managing stress: Emotions and power at work. London: SAGE.
Peeters, M. C. W., & Le Blanc, P. M. (2001). Towards a match between job demands and
sources of social support: A study among oncology care providers. European Journal of
Work and Organizational Psychology,10(1), 53–72.
Peiro
´, J. M. (2008). Stress and coping at work: New research trends and their implications
for practice. In K. Na
¨swall, J. Hellgren, & M. Sverke (Eds), The individual in the changing
working life (pp. 284–310). New York: Cambridge University Press.
Petterson, I.-L., Hertting, A., Hagberg, L., & Theorell, T. (2005). Are trends in work and
health conditions interrelated? A study of Swedish hospital employees in the 1990s.
Journal of Occupational Health Psychology,10(2), 110–120.
Pienaar, J. (2008). Skeleton key or siren song: Is coping the answer to balancing work and
well being? In K. Na
¨swall, J. Hellgren, & M. Sverke (Eds), The individual in the changing
working life (pp. 235–257). New York: Cambridge University Press.
Podsakoff, P. M., MacKenzie, S. B., Lee, J., & Podsakoff, N. P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended rem-
edies. Journal of Applied Psychology,88(5), 879–903.
Rissler, A., & Elgerot, A. (1978). Stress reactions during overtime at work. Longitudinal study
of psychological reactions during and after work. (Report No. 23). Stockholm: University
of Stockholm, Department of Psychology.
Rissler, A., & Jacobson, L. (1987). Cognitive efficiency during high workload in a final
system testing of a large computer system. In H. J. Bullinger, & B. Shackel (Eds),
Human computer interaction (Interact ’87). Amsterdam: Elsevier.
Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend?
Organizational Research Methods,9(2), 221–232.
Stanley, N., Manthorpe, J., & White, M. (2007). Depression in the profession: Social work-
ers’ experiences and perceptions. British Journal of Social Work,37(2), 281–298.
Tham, P. (2007). Why are they leaving? Factors affecting intention to leave among social
workers in child welfare. British Journal of Social Work,37(7), 1225–1246.
Astvik and Melin 359
Tham, P., & Meagher, G. (2008). Working in human services: How do experiences and
working conditions in child welfare social work compare? British Journal of Social Work,
Advance Access published 11 March, doi:10.1093/bjsw/bcm170.
Viswesvaran, C., Sanches, J. I., & Fisher, J. (1999). The role of social support in the process
of work stress: A meta-analysis. Journal of Vocational Behavior,54(2), 314–334.
von Thiele, U., Lindfors, P., & Lundberg, U. (2006). Self-rated recovery from work stress
and allostatic load in women. Journal of Psychosomatic Research,61(2), 237–242.
360 Journal of Social Work 13(4)
