Embed Size (px)
DESCRIPTION
paper
Citation preview
This article was downloaded by: [viviana puebla]On: 15 November 2014, At: 04:38Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Anxiety, Stress, & Coping: AnInternational JournalPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/gasc20
Effects of occupational stress,job characteristics, coping, andattributional style on the mental healthand job satisfaction of universityemployeesGeorge Mark a & Andrew P. Smith aa Centre for Occupational and Health Psychology , School ofPsychology, Cardiff University , Cardiff, UKPublished online: 25 Jan 2011.
To cite this article: George Mark & Andrew P. Smith (2012) Effects of occupational stress, jobcharacteristics, coping, and attributional style on the mental health and job satisfaction ofuniversity employees, Anxiety, Stress, & Coping: An International Journal, 25:1, 63-78, DOI:10.1080/10615806.2010.548088
To link to this article: http://dx.doi.org/10.1080/10615806.2010.548088
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Effects of occupational stress, job characteristics, coping, andattributional style on the mental health and job satisfaction of university
employees
George Mark and Andrew P. Smith*
Centre for Occupational and Health Psychology, School of Psychology, Cardiff University,Cardiff, UK
(Received 18 May 2010; final version received 10 December 2010)
Well-being at work has been shown to be influenced by job characteristics andindividual differences in coping styles. This study investigated the relationshipsbetween job demands, control, social support, efforts, rewards, coping, andattributional style in predicting anxiety, depression, and job satisfaction in asample of 307 university employees from the UK. Results were compared to thosefrom a sample of 120 members of the general population. Workplace demands,intrinsic and extrinsic effort, and negative coping and attributional behaviors wereassociated with high levels of depression and anxiety and low job satisfaction inuniversity employees. Rewards, social support, job control, and positive copingand attributional behaviors were associated with lower levels of depression andanxiety and high job satisfaction. The study adds to the growing research onuniversity samples by showing that a transactional approach should be adopted.This has implications for interventions and suggests that rather than just trying tochange job characteristics one should identify at-risk individuals in thispopulation and help them adopt appropriate positive coping styles.
Keywords: stress; job characteristics; coping and attributional style; jobsatisfaction; anxiety and depression; university staff
Introduction
Stress in university staff
Historically there has been a great deal of research on occupations that are typically
seen as stressful such as those with low status, control, or reward, as well other
demanding occupations such as nursing, police work, and teaching. However, there
has been less research on stress in higher education staff (Abouserie, 1996).
Abouserie states that academics willingly study other groups yet seldom study
themselves. This may be due to the perception that while academic work is not highly
paid, it is seen as highly autonomous (Winefield & Jarrett, 2001) and control is often
seen as a buffer to work stress (Karasek, 1979). However, in the last 10 years, interest
in research on academics and university employees has been growing, with
significant contributions made by Kinman (2001, 2008), Winefield (2003), and
others.
*Corresponding author. Email: [email protected]
Anxiety, Stress, & Coping
Vol. 25, No. 1, January 2012, 63�78
ISSN 1061-5806 print/1477-2205 online
# 2011 Taylor & Francis
http://dx.doi.org/10.1080/10615806.2010.548088
http://www.tandfonline.com
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Winefield and Jarrett (2001) state that stress levels in academic institutions are
high compared to many other populations, and that stress has increased significantly
over the last 15 years. Singh and Bush (1998) suggest that the persistent demands of
academic life are likely to lead to negative consequences for staff. Abouserie (1996)
states that academics have a large number of competing roles such as teaching,
research, seeking funding, writing papers, and meeting seminar and tutorial
commitments, and found that 74% of staff were moderately stressed and nearly15% were seriously stressed with lecturers the most negatively affected followed by
research assistants and professors.
Fisher (1994) claimed that academic salaries are falling and workloads increasing
and Gillespie, Walsh, Winefield, Dua, and Stough (2001), citing Association of
University Teachers (AUT) figures from 1990, stated that 49% of UK university
employees had stressful jobs. This situation may be worsening, as an AUT study in
2003 cited by Tytherleigh, Webb, Cooper, and Ricketts (2005) found that 93% of
AUT members had suffered work-related stress with high levels of dissatisfaction
with pay and workload. Gillespie et al. (2001) identify several key factors that are
commonly associated with stress in academic staff. These include work overload,
time pressure, lack of prospects, poor levels of reward and recognition, fluctuating
roles, poor management, poor resources and funding, and student interactions.
Other stressors identified from the literature by Gillespie et al. (2001) include high
expectations, low job security, lack of communication, inequality, and lack of
feedback. Indeed, a study by Kinman and Court (2010) investigated the levels of job-related stressors in a sample of 9740 academic employees at higher education
institutions in the UK including job demands, control, social support, interpersonal
relationships, role clarity, and involvement in organizational change and found that
all except one (control) exceeded the safe benchmarks as stipulated by the Health
and Safety Executive.
Winefield and Jarret (2001) report that in a sample of over 2000 Australian
university staff, 43.7% were classified as clinical cases on the General Health
Questionnaire suggesting high levels of anxiety and depression. Sharpley, Reynolds,
Acosta, and Dua (1996) found that stress was a significant problem for 25% of staff
with reports of increased anxiety, absence, injuries, illnesses, and poorer physical
health; and Blix, Cruise, Mitchell, and Blix (1994) found that 48% of staff reported
some health problems resulting from work stress. Tytherleigh et al. (2005) found
evidence that university staff exhibited significantly less organizational commitment
compared to other private and public sector workers, as well as being more stressed
by lack of control and resources, and worries about low pay and benefits.Lease (1999) states that the effect of stress in university staff is not just of
consequence to employees themselves, but can have serious consequences for
students as well. Indeed, Blix et al. (1994) report that 84% of their sample of 400
university staff reported that their productivity at work had been negatively affected
by stress and 33% felt it suffered at least 50% of the time. Boyd and Wylie (1994, cited
in Gillespie et al., 2001) found that workload and stress resulted in less time spent on
research, publishing and development, and lower teaching standards, as well as
having negative effects on staff relationships, emotional health, family relationships,
and leisure activities. Blix et al. (1994) showed that job stress significantly increased
the likelihood of staff intending to leave academia. Finally, Bowen and Schuster
(1985, cited in Gillespie et al., 2001) also reported that stress had a negative impact
64 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
on staff morale and many of the interviewed academics were angry, embittered, and
felt devalued and abandoned.
Theories of occupational stress
Two of the most popular and influential theories of workplace stress identify the
following stressors as key factors in the onset of stress-related illness. The first, the
Demand-Control-Support model (DCS; Karasek & Theorell, 1990), predicts that
high levels of job demands (external pressures and workload), low levels of job
control (over events, and chance to use skills), and low levels of social support (from
supervisors, colleagues, feedback) are associated strongly with negative health
outcomes. The model predicts interactions between demands and control, anddemands and social support so that control and support will buffer the negative
effect of job demands on health outcomes.
The second popular model, the Effort-Reward imbalance model (ERI; Siegrist,
1996) predicts that high levels of extrinsic effort (from external pressures) and
intrinsic effort (internal motivations/work ‘‘over commitment’’) and low levels of
reward (pay, job security, recognition, and promotion prospects) will significantly
predict negative health outcomes. Reward is predicted to buffer against the negative
effect of efforts on health outcomes.These two models have been found to be good predictors of physical and
psychological health outcomes including heart disease, mortality, and depression
(Van Der Doef & Maes, 1999) in many occupational groups. These two models are,
therefore, suitable for studying many of the kinds of stressors that university
employees are exposed to. Kinman and Court (2010) have investigated factors from
the demand-control-support model and other HSE risk factors and found high levels
of workload and job demands, low peer support, and poor working relationships in
university samples. This suggests that these populations may be at high risk fromstress-related illness. Also, Kinman and Jones (2008a) found that high efforts and
low rewards from the ERI model significantly predicted strain outcomes including
psychological distress, physical symptoms, and job satisfaction. Kinman and Jones
(2008b) also found that ERI factors strongly predicted work�life conflict in
university samples and found interactions between efforts and rewards in predicting
work�life conflict.
While it is clear that the study of stressful job characteristics may be helpful in the
prediction of outcomes in university employees, it is also important to take accountof how workplace stressors affect different individuals (Perrewe & Zellars, 1999). The
DCS and (to a lesser extent) the ERI model, while showing good predictive validity
for many health outcomes, do not take so much account of how different individuals
can cope or deal with the same event in different ways to result in different outcomes.
Other theoretical viewpoints however, such as many transactional stress models
(Folkman, Lazarus, Gruen, & DeLongis, 1986), emphasize the importance of coping
differences between individuals as well as other differences that may affect ways that
people appraise or assess potential stressors.Coping behaviors (adaptive or maladaptive) are seen as any cognitive or
behavioral actions used to try and reduce or tolerate environmental demands that
are seen as taxing or exceeding personal resources (Folkman et al., 1986). Coping
behaviors are central to transactional stress models and there is strong evidence that
Anxiety, Stress, & Coping 65
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
certain ‘‘positive’’ coping behaviors such as problem solving or seeking advice have
positive associations with health outcomes, and ‘‘negative’’ behaviors such as self-
blame, escape/avoidance, and wishful thinking have negative associations (Cox &
Ferguson, 1991). While coping research is very popular in the general work-stressliterature, there remain only a few studies that investigate coping in university
populations and these often report conflicting results (for example, Abouserie, 1996;
Lease, 1999). However, Kinman (2008) recently investigated the role played by Sense
of Coherence (SOC) � a generalized coping and resilience resource � in coping with
environmental stressors in academic university employees, and found that those with
a more resilient SOC exhibited significantly lower levels of negative physical and
psychological symptoms.
Attributional style (Sweeney, Anderson, & Bailey, 1986) is another individualcharacteristic related to health outcomes in clinical literature. The concept refers to
how individuals explain or perceive the causes of events on a series of dimensions
(internal or external, stable or unstable, global or local). Positive ways of attributing
are those where positive events are seen as being due to internal causes, are stable
over time, and will occur in other domains (and opposite for negative events) and
negative attributions where positive events are seen as externally caused, unstable,
and local (and the opposite for negative events). This concept is similar in some
aspects to the appraisal stage of transactional models where causes, threats, andconsequences of negative events are analyzed. However, use of this concept in
occupational stress literature is rare and even rarer in university samples.
The present study
The job characteristic variables of demands, control, social support, rewards, and
extrinsic efforts, as well as intrinsic efforts, ways of coping, and attributional style,
and their associations with self-rated depression, job satisfaction, and anxiety wereinvestigated in a university population. The study compared the relative importance
of each factor in accounting for levels of outcome variables and examined whether
there were any interactions. The same measures were used in a sample of the general
population to compare the two samples on the above variables.
The DCS and ERI variables were selected for their strength in associating with
health outcomes in occupational literature and coping for its centrality in
transactional stress models and the clinical literature. This study also differs from
many previous pieces of research in the literature in that it includes the variable ofattributional style, which is considered important due to its novelty in occupational
research and strong conceptual place in the clinical depression literature.
The present study was designed to extend research on stress in university
employees. Other recent research has also developed the area and provided a more
detailed profile than one just measuring stress levels. For example, Jacobs,
Tytherleigh, Webb, and Cooper (2007) examined relationships between stress levels,
organizational commitment, health, and performance. Stressors had a negative linear
relationship with performance measures but this relationship was modified byphysical health, psychological well-being, and organizational commitment. More
recent research by this group (Jacobs, Tytherleigh, Webb, & Cooper, 2010) found that
female university employees working in gender-incongruent roles reported higher
levels of stress whereas men who worked in gender-incongruent roles reported the
66 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
lowest levels of stress. Catano et al. (2010) also investigated factors associated with
stress in university employees. Less secure employment status and work�life
imbalance were the strongest predictors of job dissatisfaction. Psychological distress
was also predicted by work�life imbalance. The present study focused on potentially
stressful job characteristics, coping, and attributions. As such, it represents a new
approach to stress in university staff, although some similar research has been
conducted with other samples.
Seven key hypotheses were tested in the present study. The first hypothesis
predicted that university staff would report higher levels of depression, anxiety, lower
job satisfaction, and different patterns of job characteristics, coping, and attributions
compared to members of a general population sample. The second hypothesis
predicted that in university staff, ‘‘positive’’ coping (problem-focused coping, seeking
advice) would be associated with low levels of depression and anxiety and high job
satisfaction, and that ‘‘negative’’ coping (self-blame, escape/avoidance, wishful
thinking) would be associated with high levels of anxiety and depression and low
job satisfaction. The third hypothesis predicted that positive attributional-style
behaviors would be associated with low levels of anxiety, depression, and high job
satisfaction, and negative attributions would be associated with high anxiety,
depression, and low job satisfaction. The fourth hypothesis predicted that high job
demands would be associated with high depression, anxiety, and low satisfaction,
and high levels of decision authority, skill discretion, and social support would be
associated with low levels of anxiety, depression, and high job satisfaction. It was
also predicted that control and social support would interact with demands to
produce a buffering effect in reporting of anxiety, depression, and job satisfaction.
The fifth hypothesis predicted that high extrinsic and intrinsic effort would be
significantly associated with high depression, anxiety, and low satisfaction, and high
levels of intrinsic reward would be associated with low levels of anxiety, depression,
and high job satisfaction. It was also predicted that rewards would significantly
interact with intrinsic and extrinsic effort to buffer their effects in reporting of
anxiety, depression, and job satisfaction. The sixth hypothesis predicted that positive
attributions and problem-focused coping would significantly interact with the
negative job characteristics of high job demands and extrinsic efforts, so that
positive individual characteristics would buffer the negative association between
negative job characteristics and anxiety, depression, and job satisfaction. The final
hypothesis tested was that coping and attributional style, intrinsic and extrinsic
effort, rewards, demands, skill discretion, decision authority, and social support
would account for a significant amount of the variance in anxiety, depression, and
job satisfaction in university employees. It was also predicted that the addition of
coping and attributional style would significantly increase the explained variance in
anxiety, depression, and satisfaction over the use of DCS and ERI variables alone.
Method
Participants
The participants were a sample of 307 university employees and 120 members of the
general population. A bulk email was sent to all academic and administrative staff
(approximately 2800) at Cardiff University in the UK requesting participants for a
Anxiety, Stress, & Coping 67
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
study on stress and health at work. Those who responded were sent further details
about the aims and methodology and a questionnaire pack with freepost return
envelope. The general public participants were contacted from a pool of potential
participants (N�200) who had previously taken part in psychological research at theuniversity and had indicated that they would be willing to participant in future
research. Response rates for the university and general population samples were 11%
and 60%, respectively.
An a priori power analysis (Gpower; Buchner, Faul, & Erdfelder, 1992) showed
that at a significance level of .05, 307 participants would be sufficient for a medium
effect size of approximately .28 for an experimental power of .80.
Measures
A 24-page questionnaire booklet was produced, which contained an instruction page
informing participants about the purpose of the experiment, their right to withdraw,
and the anonymous treatment of data (i.e., that no personally identifiable data would
be collected and data would be securely stored). The questionnaire pack contained
demographic questions as well as the following questionnaires.
The Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983)
was used to measure mental health in the forms of anxiety and depression. It is a 14-item self-report scale with two 7-item subscales that measure anxiety and depression.
Fourteen mood-related descriptions are presented (e.g., ‘‘I feel cheerful’’), and
participants respond on a 4-point Likert scale on how often they have experienced
the suggested moods in the past week. All items were positively coded with responses
ranging from ‘‘not at all’’ (0) to ‘‘nearly all the time’’ (3) with scores summed for each
subscale with a maximum of 21 and higher scores indicating more anxiety or
depression. A score of 11 or more indicated possible clinical anxiety or depression
(Zigmond & Snaith, 1983). Cronbach’s a scores were calculated as .87 for anxiety and.80 for depression subscales.
Coping was measured using the revised version of the Ways of Coping Checklist
(WCCL; Folkman & Lazarus, 1980). The 42 items from Vitaliano, Russo, Carr,
Maiuro, and Becker’s (1985) revised version of the Ways of Coping Checklist were
used. Five subscales were used to measure each of the following types of coping
behavior: problem-focused coping (e.g., identifying problems and seeking solu-
tions), seek advice (e.g., speaking to peers/others for support), self-blame (blaming
oneself for negative events), wishful thinking (e.g., wishing for problems todisappear), and escape/avoidance (ignoring or avoiding problem). Participants
were asked to think of a recent stressful experience at work and to indicate on the
4-point Likert scale (0�3) how often they used each of the suggested coping
behaviors, from across all five subscales. Mean scores for each subscale were
calculated and were converted to percentages of the maximum (due to different
items in the subscales) with a higher score indicating a greater propensity to use
that type of coping (i.e., positively coded). Cronbach’s a scores ranged from .73 to
.88 for all subscales.A 21-item version of the ERI Questionnaire was used (as by Kuper, Singh-
Manoux, Siegrist, & Marmot, 2002) to measure intrinsic effort (internal motivation
or ‘‘over commitment’’ to work), extrinsic effort (from external demands), and
internal reward (e.g., perceptions that rewards and recognition are adequate).
68 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Descriptions were given of various workplace circumstances (e.g., ‘‘My job security is
poor’’). Participants were asked to indicate if they had experienced the situation, and
if so, to respond on a 4-point Likert scale indicating how distressing they find each,
from ‘‘not at all’’ to ‘‘very distressing.’’ Scores for each subscale were converted to
percentages with higher scores indicating higher levels of the presence of that work
situation (positively coded). Cronbach’s a scores were .78 to .87 for all subscales.
The Job Content Questionnaire (JCQ; Karasek, Brisson, Kawakami, Houtman,
Bongers, & Amick, 1988) was used to measure three subscales that measure job
demands (e.g., workload), job control (from subfactors of decision authority and
skill discretion), and levels of social support. Participants were presented with 27
suggested work characteristics/situations, and they were asked to respond on a
4-point scale as to how often they had experienced these (e.g., ‘‘Do you have to work
very fast’’). Higher scores indicate a greater presence of that category of job
conditions. Scores were positively coded for each subscale and were converted to
percentages. Cronbach’s a scores ranged from .72 to .89.
Attributional style was measured using a modified version of the Attributional
Style Questionnaire (ASQ; Peterson, 1991). Participants were presented with 12
hypothetical work situations, half of which were positive (e.g., you get a raise) and
half of which were negative (e.g., you cannot complete all of your work). Of these 12
situations, half were affiliation oriented (related to relationships at work: 3 positive, 3
negative) and half were achievement oriented (to do with work achievements or
successes/failures: 3 positive, 3 negative). The participants considered their work
situation and were asked to make an attribution about why they thought that this
situation occurred (such as ‘‘You get a raise, is this likely to be due to. . .’’).Participants responded for each event on three anchored subscales, which represent
the three attributional dimensions of: likely locus of cause (internal � it’s due to me;
external � it’s due to the situation); stability of the cause (stable �will happen again
with future events; unstable � just this event); and globality of cause (glo-
bal �happens with other types of events; local � just this type of event). Responses
were summed across situations, which gave a score for each of the attributional
dimensions of locus, stability, and globality, for both positive and negative events.
Each participant had three scores for positive events and three for negative events:
for internal/external attributions (high score � internality, low score �externality),
stable/unstable attributions (high score �stable, low score �unstable), and global/
local attributions (high �global, low � local). All items were positively coded and
reliabilities ranged from .53 to .78.
Job satisfaction was measured using the satisfaction subscale of the Copenhagen
Psychosocial Questionnaire (COPSOQ; Kristensen, Hannerz, Høgh, & Borg, 2005).
It was comprised of four items relating to satisfaction with work conditions. The job
factors related to the job as a whole, job prospects, and usage of abilities, and
participants respond to each item on a 4-point Likert scale indicating their
satisfaction level with conditions with responses from ‘‘highly unsatisfied’’ (0) to
‘‘very satisfied’’ (3). Total scores were used (maximum �12) with a higher score
indicating higher job satisfaction. Cronbach’s reliability was found to be .78.
Participants were also asked whether they had suffered an illness that they
thought was caused or made worse by work (Smith, 2001). This was measured by a
‘‘Yes’’ or ‘‘No’’ response.
Anxiety, Stress, & Coping 69
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Procedure
Those who indicated they were willing to participate were sent a questionnaire
package with a freepost return envelope. Completion time was approximately 20�30
minutes, and all instructions were given as specified or adapted from original
questionnaire authors.
The anonymous treatment of data was emphasized, and it was stated that queries
or requests for further information could be made by contacting the research team.Ethical approval was granted by Cardiff University Psychology ethics committee in
accordance with APA and BPS guidelines.
Design and analysis
This study used a cross-sectional, independent-subjects design with a university
sample and a nonmatched comparison group. Independent variables were coping
styles, levels of job demands, control, social support, efforts and rewards,
and attributional style (internality, stability, and globality attributions for positiveand negative work events). Dependent variables were levels of anxiety, depression, and
job satisfaction. Data were analyzed using multiple linear regression, correlations, and
ANOVA, with the statistical package being SPSS 15.
Results
Participant demographics
The university sample was 73.6% female with a mean age of 41.9 years (SD �10.68)
and an average working week of 38 hours. The general population sample was 68.9%female with a mean age of 45.6 years (SD �11.49) and an average working week of 35
hours. The samples also were comprised of a well-distributed range of occupational
roles (University �Administration: 21.8%, Lecturer: 25.7%, Professor: 3.9%,
Research: 18.2%, Manager: 17.9%, Other: 12.4%. General population �Administra-
Administration: 20.8%, Teacher: 11.7%, Nursing: 13.3%, Sales: 6.7%, Manager:
10.8%, Public sector: 8.3%, Other: 24.9%). Therefore, the two samples were relatively
similar in terms of gender distribution, mean age, and a broad spread of job roles.
Comparison of the university and general population groups
Results for self-rated anxiety scores on the HAD showed that 31.6% of university
staff scored over the clinical cutoff point (a score of 11; Zigmond & Snaith, 1983),
indicating that they could be potential clinical cases compared to 18.3% of the
general population. For depression, 7.8% of the university staff showed clinical
scores compared to 5.8% of the general population.
ANOVA calculationswere carried out to compare university employees and general
population samples across all variables, the results of which can be seen in Table 1.
University staff showed significantly higher levels of anxiety, depression, job demands,extrinsic and intrinsic effort, and lower levels of reward, as well as higher levels of
decision authority (control over work situations) and skill discretion (more chances to
use personal competencies). University staff were also significantly more likely to make
stable attributions for the causes of positive and negative events (akin to a ‘‘nothing
70 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
changes’’ attitude). There were no significant differences for job satisfaction, levels ofsocial support, attributed locus of control, global/local attributions, problem-focused
coping, seeking advice, self-blame, escape/avoidance, and wishful thinking.
Participants were asked to indicate whether they felt that any illness in the
past year had been caused or made worse by stress and what the illness was. Forty
percent of the university staff reported that work was related to one or more past or
current health complaints compared to 25.8% of the general population. Of all
health-related problems reported, stress-related mental health was the most common
complaint in university staff (28.5% of all complaints). Gastrointestinal problemsand musculoskeletal pain were second and third most common for university staff.
However, in the general population gastrointestinal and musculoskeletal complaints
were the most common with stress-related mental health third. All other illnesses
appeared to be equally common in the two groups.
Predictors of well-being in university staff
Table 2 shows significant correlations between independent variables and anxiety,
depression, and job satisfaction for university staff. It can be seen that so called
Table 1. ANOVAs showing significant differences between university (N�307) and general
population (N�120) samples on job characteristics and individual difference variables.
Sample Mean SD F Sig.
Anxiety University 8.36 4.36
General pop 6.91 4.07 9.90 .002
Depression University 4.95 3.46
General pop 4.04 3.49 5.88 .016
Job demands University 64.98 22.15
General pop 54.76 24.51 17.15 .001
Skill discretion University 73.28 17.16
(control) General pop 65.17 21.06 16.73 .001
Decision authority University 64.12 19.46
(control) General pop 56.74 25.04 10.44 .001
Extrinsic effort University 30.03 21.28
General pop 24.78 20.94 5.30 .022
Intrinsic effort University 21.77 22.15
General pop 13.33 18.38 13.71 .001
Reward University 81.37 20.52
General pop 88.01 13.99 10.64 .001
Stable attributions University 5.33 .80
Positive event General pop 5.13 .77 4.99 .026
Stable attributions University 4.66 1.02
Negative event General pop 4.41 .90 5.09 .025
Anxiety, Stress, & Coping 71
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
‘‘positive’’ coping, attributional behaviors, and job characteristics such as problem-
focused coping, internal and stable attributions for positive events, social support,
decision authority, skill discretion, and reward show negative correlations of small to
moderate sizes with anxiety and depression and positive correlations with satisfaction.
It can also be clearly seen that the more ‘‘negative’’ characteristics such as self-
blame, wishful thinking, escape/avoidance, internal and global attributions for
negative events, job demands, and extrinsic and intrinsic effort have positive
relationships with anxiety and depression and negative relationships with job
satisfaction. Correlations ranged from .11 to .68 and most were significant to at
least pB.01.
Multiple linear regressions were carried out to investigate the associations of
multiple independent variables with anxiety, depression, and satisfaction in the
university sample. For all regressions, models were derived by manual backward
selection so that all relevant variables were entered in the first stage, and the least
significant predictor (as determined by significance level and standard beta weight)
was removed in each following stage. This resulted in models where all final
independent variables (IVs) were significant predictors of the dependent variable
(DV). However, the simultaneous entry of so many IVs means that some individual
variables that had been significantly associated with the DV (e.g., in the correlations)
were not found to be significant in the final regression models (e.g., escape/avoidance
in the anxiety regression) and were thus removed. As the variance in the DV for these
IVs was accounted for by other better predictors, their removal was necessary for the
most parsimonious model.
Before the main analyses were conducted, initial models were calculated. These
were for: ways of coping and attributions against anxiety, depression, and job
Table 2. Significant correlations between anxiety, depression, satisfaction, and coping,
attributions, and job characteristics in the university sample (N�307).
Anxiety Depression Satisfaction
Problem-focused coping �.117* .164**
Self-blame .468** .342** �.230**
Wishful thinking .353** .281** �.299**
Escape/avoidance .363** .345** �.307**
Internal attributions, positive event �.169** �.228** .190**
Internal attributions, negative event .146*
Stable attributions, positive event �.157** �.120* .161**
Global attributions, negative event .129* .152* �.132*
Job demands .408** .360* �.112*
Social support �.308** �.437** .444**
Skill discretion �.118* .366**
Decision authority �.266** �.284** .435**
Extrinsic effort .476** .479** �.193**
Intrinsic effort .604** .549** �.221**
Intrinsic reward �.332** �.395** .682**
*pB.05. **pB.01.
72 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
satisfaction; job demands, control, and social support (from the work of Karasek),
and efforts and rewards (from the work of Siegrist) against anxiety, depression, and
job satisfaction. These results are briefly summarized here before description of the
main models (where all variables were simultaneously entered).
Self-blame, escape/avoidance, and global attributions for negative events, all had
positive relationships with anxiety and global attributions for positive events had anegative relationship with anxiety. Self-blame was the most important factor by
standardized beta weight. Self-blame and escape/avoidance had positive relationships
with depression, and internal attributions for positive events showed a negative
relationship with depression. Problem-focused coping and global attributions for
positive events were associated with an increase in job satisfaction, and wishful
thinking, escape/avoidance, global attributions for negative events, and seeking
advice were all associated with lower job satisfaction.
The variables of job demands, control and support, and efforts and reward were
then entered against anxiety, depression, and job satisfaction. For anxiety, job
demands and intrinsic effort were associated with significantly higher scores and
social support and decision authority with lower anxiety. Social support, skill
discretion, and intrinsic reward were significantly associated with lower depression
and job demands, intrinsic effort, and extrinsic effort associated with significantly
higher depression scores. Finally, high job demands were the only factor that
significantly associated with lower job satisfaction, but higher social support, skill
discretion, decision authority, and intrinsic rewards all predicted significantly higher
job satisfaction. There was little evidence of interactions between the jobcharacteristics. One exception was in the analysis of depression, where high decision
authority acted as a buffer against the effects of high demands.
Table 3 shows the three final regressions where all variables were entered and
selected by manual backward selection. Intercorrelations of independent variables
showed none over .80, suggesting no multicollinearity. When all job characteristics and
individual difference variables, age, and sex were regressed against anxiety, depression,
and job satisfaction, the following was found. Self-blame, wishful thinking, job
demands, and intrinsic effort were significantly associated with increased anxiety
scores, and problem-focused coping, social support, and decision authority with lower
anxiety scores. Intrinsic effort was the strongest predictor by standardized beta weight,
followed by job demands and self-blame, with all other factors showing similar
standardized betas. These factors accounted for 55% of the variance in anxiety.
For depression, escape/avoidance, job demands, and extrinsic and intrinsic effort
related to a significant increase in score, and internal attributions for positive events,
social support, skill discretion, and intrinsic rewards associated with significantly
lower depression. Intrinsic effort and social support had the strongest associationswith depression by standardized beta, followed by escape/avoidance and extrinsic
effort. The above variables accounted for 52% of the variance in depression scores.
Finally, seeking advice, escape/avoidance, and global attributions for negative
events significantly associated with lower job satisfaction, and social support, skill
discretion, decision authority, global attributions for positive events, and intrinsic
reward significantly associated with increases in job satisfaction. These factors
accounted for 57% of the variance in satisfaction. Reward was by far the most
important factor, followed by social support and global-negative attributions, with
other factors all of similar standardized beta weights.
Anxiety, Stress, & Coping 73
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Discussion
Six of the seven hypotheses presented in this paper were partially or fully supported
showing strong relationships between outcomes, coping, attributional style, job
characteristics, and efforts in university employees. The results confirmed predictions
from the established theories of occupational stress described in the introduction. In
addition, they generally confirmed prior research on stress in university staff. Data
showed that there were some significant differences between university and general
Table 3. All job characteristics and individual differences regressed against anxiety,
depression, and satisfaction in university staff (N�307).
Anxiety Beta weight Std. error
Standardized
beta weight t Sig.
(Constant) 5.742 1.170 4.908 .001
Problem-focused coping �.920 .447 �.096 �2.061 .040
Seek advice .668 .354 .087 1.887 .060
Self-blame 1.086 .356 .174 3.051 .003
Wishful thinking .764 .378 .109 2.019 .045
Job demands .039 .009 .197 4.170 .001
Social support �.024 .009 �.129 �2.797 .006
Decision authority �.021 .010 �.094 �2.049 .041
Intrinsic effort/over commitment .080 .010 .415 8.135 .001
Model: R�.742, R2�.551 F: 39.61 .001
Depression Beta weight Std. error
Standardized
beta weight t Sig.
(Constant) 8.579 1.138 7.539 .001
Escape/avoidance .848 .256 .150 3.314 .001
Internal attribution: positive event �.386 .161 �.103 �2.400 .017
Job demands .019 .009 .120 2.167 .031
Social support �.035 .007 �.245 �5.044 .001
Skill discretion �.020 .010 �.101 �2.093 .037
Extrinsic effort .022 .011 .138 2.097 .037
Intrinsic effort/over commitment .042 .009 .276 4.680 .001
Intrinsic rewards �.019 .008 �.113 �2.298 .022
Model: R�.719, R2�.517 F: 37.41 .001
Job satisfaction Beta weight Std. error
Standardized
beta weight t Sig.
(Constant) �.693 .839 �.826 .410
Seek advice �.350 .180 �.083 �1.949 .052
Escape/avoidance �.448 .170 �.114 �2.635 .009
Global attribution: positive event .258 .121 .118 2.128 .034
Global attribution: negative event �.325 .127 �.158 �2.565 .011
Social support .015 .005 .142 2.886 .004
Skill discretion .015 .006 .108 2.342 .020
Decision authority .013 .006 .102 2.092 .037
Intrinsic reward .060 .006 .512 10.200 .001
Model: R�.754, R2�.569 F: 37.87 .001
74 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
population samples, with university employees reporting more anxiety and depres-
sion (even up to clinical levels). It was also found that university staff were more
likely to claim that workplace conditions had caused or made an illness worse and
were twice as likely to complain of stress or anxiety and depression-related illness.
While these groups were sampled in different ways, they were similar in terms of
demographic factors and distribution of job types. It was also shown that the groups
did not differ significantly in coping behaviors or attributional behaviors. Thissuggests the possibility that the difference in work-related mental health outcomes
may be due to differential exposure to stressors rather than differences in
susceptibility to stress, especially as university staff had more control at work but
had more job demands and felt less rewarded.
While these data were cross-sectional and not causal, they provide a theoretical
basis for organizational interventions, as they help to show which independent
variables are most strongly associated with outcomes. Possible interventions based on
these results could include: provision of improved social support systems, training in
problem-focused coping methods and discouragement of negative coping such as
self-blame and escape avoidance, enhancement of reward systems, skills training for
the enhancement of decision authority and skill discretion, or awareness training
on the dangers of becoming ‘‘overcommitted’’ to work and having negative beliefs
about the causes of events.
Such multimodal interventions could not be recommended on the basis of jobcharacteristics research alone (even if those characteristics are subjectively reported),
because such research can only recommend empirically based interventions based on
job characteristics factors. The information from this study could also be particularly
useful in situations where resources for intervention are limited, and only variables
most strongly related to outcomes can be targeted, or where certain types of
intervention are not feasible (e.g., when reduction of job demands or enhancement of
control is constrained) but where other more individual factors may be more
amenable to change.
The methodology of this study had some limitations. One problem was the low
response rate. This is not unique to the present study and other research in this area
has obtained lower than expected response rates that may reflect the nature of the
topic under investigation. Indeed, in some studies questionnaires were sent to home
addresses to capture highly stressed individuals who may have been off sick
(Tytherleigh et al., 2005; Tytherleigh, Jacobs, Webb, & Ricketts, 2007). The study
was cross-sectional and, therefore, cause and effect conclusions about relationships
between variables cannot be made, and as the results were self-report they may havebeen open to biases from negative affect or social desirability. Finally, a large number
of comparisons were carried out, which could lead to an inflation of family wise error
rate, which was not controlled for (meaning that the results should be treated with
caution). However, one can suggest that using the Bonferroni method for controlling
family-wise error is too conservative, as it controls for the chances of any false
positives rather than just reducing the chances of individual false positives. Thus it is
extremely likely to lead to a significant number of false negatives � just as undesirable
an outcome as false positives (Scheid, 2002).
Many of the improvements that could be applied to this study are related to
correcting methodological problems, for example, using a larger and more
representative sample. Also it is recommended that future research also includes
Anxiety, Stress, & Coping 75
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
alternative methods of data collection such as qualitative measures, interviews,
supervisor ratings, absence data, or measurement of workload by some other more
objective measure. More investigation could be done to see if there were any
differences between those with different job roles, as work by Winefield and Jarret
(2001) and others suggest that different roles may give rise to differing levels of
stressors (e.g., professors, lecturers, research staff).
To enable better understanding of causal relationships between the variables, a
longitudinal method should be used, looking at the same organizational and
individual variables but with an intervention stage to see if predicted improvements
to outcomes could be found. Finally, while several of the hypotheses were derived
with transactional stress theories in mind (e.g., the moderating role of coping on
outcomes), the method did not truly reflect the transactional theory. A method that
more accurately tested transactional models could be used (e.g., by examining the
relationships between objective job characteristics and subjective perceptions of
stress as mediated by individual differences, and then examining the relationships
between subjective perceptions and health outcomes as moderated by individual
differences).
In conclusion, the data presented here add to the growing literature that shows
that university staff may suffer from high levels of anxiety, depression, and stress-
related illness compared to general population samples. Our results show that there
are strong associations between the traditional variables of efforts, demands, control,
supports, and rewards, and depression, anxiety, and job satisfaction and also
between coping and attributional style and these outcomes. This is important given
the failure of traditional models such as DCS and ERI to take account of individual
differences, and the paucity of research on attributional style in work-related stress.
The fact that both environmental and individual characteristics were important in
the prediction of outcomes supports the premise of transactional stress theories and
illustrates the need for models and research that integrates individual and job
characteristics factors (see Mark & Smith, 2008). The inclusion of attributional style
in this study is strength, and the fact that the variable has received little attention in
occupational stress literature is curious, particularly given its importance in the
depression literature. Further exploration of the importance of this variable may
therefore remain a good avenue for further research.
Further research needs to be conducted on university staff as workplace stress is
clearly a serious problem in this population, with 32% with clinical anxiety scores on
the HAD, 8% with clinical depression scores, and 40% claiming that work stress had
negatively affected their health. This is especially important with regards to the
potential consequences for students. Workplace demands, intrinsic and extrinsic
effort, and negative coping and attributional behaviors were associated with high
levels of depression and anxiety and low job satisfaction in university employees.
Rewards, social support, job control, and positive coping and attributional behaviors
were associated with lower levels of depression and anxiety and high job satisfaction.
Continuing research into stress, health, and satisfaction in university staff could help
to provide empirical evidence to form the basis for interventions that would benefit
universities, employees, and students. The results from the present study have
identified job characteristics and personal resources that could be important in
further research.
76 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
References
Abouserie, R. (1996). Stress, coping, strategies and job satisfaction in university academicstaff. Educational Psychology, 16(1), 49�56. doi: 10.1080/0144341960160104
Blix, A.G., Cruise, R.J., Mitchell, B.M., & Blix, G.G. (1994). Occupational stress amonguniversity teachers. Educational Research, 36(2), 157�169.
Buchner, A., Faul, F., & Erdfelder, E. (1992). GPOWER: A priori-, post hoc-, andcompromise power analyses for the PC [computer program]. Bonn, Germany: BonnUniversity.
Catano, V., Francis, L., Haines, T., Kirplani, H., Shannon, H., Stringer, B., & Lozanzki, L.(2010). Occupational stress in Canadian universities: A national survey. InternationalJournal of Stress Management, 17(3), 232�258.
Cox, T., & Ferguson, E. (1991). Individual differences, stress and coping. In C.L. Cooper &R. Payne (Eds.), Personality and stress: Individual differences in the stress process. New York:Wiley.
Fisher, S. (1994). Stress in academic life. Buckingham: The Society for Research into HigherEducation and Open University Press.
Folkman, S., & Lazarus, R.S. (1980). An analysis of coping in a middle-aged communitysample. Journal of Health and Social Behavior, 21, 219�239. doi: 10.1037/0022-3514.50.3.571
Folkman, S., Lazarus, R.S., Gruen, R.J., & DeLongis, A. (1986). Appraisal, coping, healthstatus, and psychological symptoms. Journal of Personality and Social Psychology, 50(3),571�579. doi: 10.1037/0022-3514.50.3.571
Gillespie, N.A., Walsh, M., Winefield, A.H., Dua, J., & Stough, C. (2001). Occupational stressin universities: Staff perceptions of the causes, consequences and moderators of stress. Work& Stress, 15(1), 53�72.
Jacobs, P.A., Tytherleigh, M.Y., Webb, C., & Cooper, C.L. (2007). Predictors of workperformance among higher education employees: An examination using the ASSET modelof stress. International Journal of Stress Management, 14(2), 199�210.
Jacobs, P.A., Tytherleigh, M.Y., Webb, C., & Cooper, C.L. (2010). Breaking the mold: Theimpact of working in gender-congruent versus gender-incongruent role on self-reportedsources of stress, organizational commitment, and health in UK universities. InternationalJournal of Stress Management, 17(1), 21�37.
Karasek, R. (1979). Job demands, job decision latitude and mental strain: Implications for jobredesign. Administrative Science Quarterly, 24, 285�306. doi: 10.2307/2392498
Karasek, R., Brisson, C., Kawakami, N., Houtman, I., Bongers, P., & Amick, B. (1988). TheJob Content Questionnaire (JCQ): An instrument for internationally comparative assess-ments of psychosocial job characteristics. Journal of Occupational Health Psychology, 3(4),322�355. doi: 10.1037/1076-8998.3.4.322
Karasek, R., & Theorell, T. (1990). Healthy work: Stress, productivity and the reconstruction ofworking life. New York: Basic Books.
Kinman, G. (2001). Pressure points: Review of research on stressors and strains in UKacademics, Educational Psychology, 21, 473�492. doi: 10.1080/01443410120090849
Kinman, G. (2008). Work stressors, health and sense of coherence in UK academic employees.Educational Psychology, 28, 823�835. doi: 10.1080/01443410802366298
Kinman, G., & Court, S. (2010). Psychosocial hazards in UK universities: adopting a riskassessment approach. Higher Educational Quarterly, 64, 413�428. doi: 10.1111/j.1468-2273.2009.00447.x
Kinman, G., & Jones, F. (2008a). Effort-Reward imbalance and over-commitment: predictingstrain in academic employees in the UK. International Journal of Stress Management, 15,381�395. doi: 10.1037/a0013213.
Kinman, G., & Jones, F. (2008b). Effort-Reward imbalance, over-commitment and work-lifeconflict: testing an expanded model. Journal of Managerial Psychology, 23, 236�251. doi:10.1108/02683940810861365.
Kristensen, T.S., Hannerz, H., Høgh, A., & Borg, V. (2005). The Copenhagen PsychosocialQuestionnaire � A tool for the assessment and improvement of the psychosocial workenvironment. Scandinavian Journal of Work and Environmental Health, 31(6), 438�449.
Anxiety, Stress, & Coping 77
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
Kuper, H., Singh-Manoux, A., Siegrist, J., & Marmot, M. (2002). When reciprocity fails:Effort-reward imbalance in relation to coronary heart disease and health functioning withinthe Whitehall II Study. Occupational and Environmental Medicine, 59, 777�784. doi:10.1136/oem.59.11.777
Lease, S.H. (1999). Occupational role stressors, coping, support, and hardiness as predictors ofstrain in academic faculty: An emphasis on new and female faculty. Research in HigherEducation, 40(3), 285�307. doi: 10.1023/A:1018747000082
Mark, G.M., & Smith, A.P. (2008). Stress models: A review and suggested new direction. InJ. Houdmont & S. Leka (Eds.), EA-OHP series (Vol. 3, pp. 111�144). Nottingham:Nottingham University Press.
Perrewe, P.L., & Zellars, K.L. (1999). An examination of attributions and emotions in thetransactional approach to the organizational stress process. Journal of OrganizationalBehavior, 20, 739�752.
Peterson, C. (1991). On shortening the expanded attributional style questionnaire. Journal ofPersonality Assessment, 56, 179�183. doi: 10.1207/s15327752jpa5601_16
Scheid, S. (2002). Multiple testing: False discovery rate and the q-value. Berlin: Max-Planck-Institute for Molecular Genetics. Computational Diagnostics.
Sharpley, C.F., Reynolds, R., Acosta, A., & Dua, J.K. (1996). The presence, nature and effectsof job stress on physical and psychological health at a large Australian University. Journal ofEducational Administration, 34, 73�86. doi: 10.1108/09578239610128630
Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal ofOccupational Health Psychology, 1, 27�41. doi: 10.1037/1076-8998.1.1.27
Singh, S.N., & Bush, R.F. (1998). Research burnout in tenured marketing professors: Anempirical investigation. Journal of Marketing Education, 20(1), 4�15. doi: 10.1177/027347539802000102
Smith, A.P. (2001). Perceptions of stress at work. Human Resource Management Journal, 11,74�86.
Sweeney, P., Anderson, K., & Bailey, S. (1986). Attributional style in depression: A meta-analytic review. Journal of Personality and Social Psychology, 50, 974�991. doi: 10.1037/0022-3514.50.5.974
Tytherleigh, M.Y., Jacobs, P.A., Webb, C., & Ricketts, C. (2007). Gender, health and stress inEnglish university staff � Exposure or vulnerability? Applied Psychology: An InternationalReview, 56(2), 267�287. doi: 10.1111/j.1464-0597.2006.00254
Tytherleigh, M.Y., Webb, C., Cooper, C.L., & Ricketts, C. (2005). Occupational stress in UKhigher education institutions: A comparative study of all staff categories. Higher EducationResearch & Development, 24(1), 41�61. doi: 10.1080/0729436052000318569
Van Der Doef, M., & Maes, S. (1999). The Job-Demand (-Support) Model andpsychological well-being: A review of 20 years of empirical research. Work & Stress,13(2), 87�114.
Vitaliano, P.P., Russo, J., Carr, J.E., Maiuro, R.D., & Becker, J. (1985). The ways of copingChecklist psychometric properties. Multivariate Behavioral Research, 20, 3�26. doi: 10.1207/s15327906mbr2001_1
Winefield, A.H. (2003). Stress in university academics. In M.F. Dollard, A.H. Winefield, &H.R. Winefield (Eds.), Occupational stress in the service professions. London: Taylor &Francis.
Winefield, A.H., & Jarrett, R. (2001). Occupational stress in university staff. InternationalJournal of Stress Management, 8(4), 285�298. doi: 10.1023/A:1017513615819
Zigmond, A.S., & Snaith, R.P. (1983). The hospital anxiety and depression scale. ActaPsychiatrica Scandinavia, 67, 361�370. doi: 10.1111/j.1600-0447.1983.tb09716
78 G. Mark and A.P. Smith
Dow
nloa
ded
by [
vivi
ana
pueb
la]
at 0
4:38
15
Nov
embe
r 20
14
