Scandinavian Journal of Psychology, 2002, 43, 397 405

Relationships between exposure to bullying at work and psychological

Blackwell Publishing Ltd

and psychosomatic health complaints: The role of state negative

affectivity and generalized self-efficacy
EVA GEMZE MIKKELSEN1 and STLE EINARSEN2
1
Department of Psychology, University of Aarhus, Denmark
2
Department of Psychosocial Science, University of Bergen, Norway

Mikkelsen, E. G. & Einarsen, S. (2002). Relationships between exposure to bullying at work and psychological and psychosomatic health
complaints: The role of state negative affectivity and generalized self-efficacy. Scandinavian Journal of Psychology, 43, 397 405.
The aim of this study was to investigate the relationships between exposure to behaviors identified as workplace bullying and self-report
measurements of psychological and psychosomatic health complaints. Secondly, we investigated whether these relationships were mediated by
the state negative affectivity of the victim. Thirdly, we investigated the extent to which generalized self-efficacy moderated the relationships
between exposure to bullying behaviors and health complaints. Two hundred and twenty-four white- and blue-collar employees from a Danish
manufacturing company participated in the study. Exposure to bullying behaviors was associated with an increase in psychological health
complaints, increased levels of psychosomatic complaints and an elevated level of state negative affectivity. Whereas bullying by itself accounted
for 27% of the variance in psychological health complaints and 10% of the variance in psychosomatic complaints, the results pointed to state
negative affectivity as a partial mediator of the relationships between exposure to bullying behaviors and both measures of self-reported
health. Though not directly related to reported exposure to bullying behaviors, generalized self-efficacy seemed to act as a moderator of the
relationship between exposure to bullying behaviors and psychological health complaints.
Key words: Bullying, harassment, victimization, negative affectivity, self-efficacy.
Eva Gemze Mikkelsen, Department of Psychology, University of Aarhus, Asylvej 4, DK-8240 Risskov, Denmark . E-mail: egemzoe@hotmail.com

The aim of the present study was to investigate the relation-
ships between exposure to bullying behaviors at work and
psychological and psychosomatic health complaints, and the
role of state negative affectivity as mediator and generalized
self-efficacy as a moderator of these relationships. Studies
on organizational stress (Chen & Spector, 1991; Keenan &
Newton, 1984) have shown that interpersonal conflicts and
aggression constitute an important work stressor adversely
affecting employee health and wellbeing. Concurrently, the
last decade has witnessed a growth in studies focusing on
systematic interpersonal harassment and bullying in the
workplace (see Einarsen, 2000; Hoel, Rayner & Cooper,
1999, for reviews). The concept of workplace bullying, some-
times referred to as workplace abuse ( Keashly, 1998), mob-
bing ( Leymann, 1996) or workplace harassment (Brodsky,
1976), refers to situations where a person repeatedly and
over a period of time is exposed to negative acts (i.e. con-
stant abuse, offensive remarks or teasing, ridicule or social
exclusion) on the part of co-workers, supervisors or subor-
dinates ( Einarsen, 2000). A defining characteristic of bully-
ing is prolonged exposure to repeated negative acts. Contrary
to isolated interpersonal conflicts in organizations, bullying
is systematic aggression aimed at one or more individuals by
another individual or a group ( Hoel et al., 1999). The victim
will typically experience these acts as extremely offensive,
degrading, and unfair. The feeling of being unjustly treated
is central to most victims of bullying. In addition, bullying
appears not to be an eitheror phenomenon, but rather an
escalating process, where the victim gradually faces increas-
ingly frequent as well as increasingly intense aggressive
behaviors (Einarsen, 2000). Furthermore, the behaviors
involved seem mainly to be of a verbal and nonphysical
nature, including indirect behaviors such as shunning.
A number of studies have shown significant relation-
ships between exposure to bullying and both victim and co-
workers health and wellbeing (Brodsky, 1976; Einarsen,
Matthiesen & Skogstad, 1998; Zapf, Knorz & Kulla, 1996).
In a study of male industrial workers, exposure to workplace
bullying explained 23% of the variance in psychological
health and wellbeing (Einarsen & Raknes, 1997). On the
whole, victims exhibit a wide range of psychological, psycho-
somatic and musculoskeletal symptoms, such as low self-
esteem, anxiety, depression, concentration difficulties, chronic
fatigue, sleep problems, stomach problems, back- and head-
aches, anger, self-hatred and suicidal thoughts (Bjrkqvist,
sterman & Hjelt-Bck, 1994; Brodsky, 1976; Einarsen,
Raknes, Matthiesen & Hellesy, 1996; Leymann, 1990). In
an interview study of 17 victims of bullying in a Finnish
university, Bjrkqvist et al. (1994) found that all victims
reported insomnia, various nervous symptoms, melancholy,
apathy, lack of concentration and socio-phobia. Clinical
observations have also shown other symptoms related to
exposure to workplace bullying, such as social isolation,
social maladjustment, psychosomatic illnesses, depression,
helplessness, anger, compulsions, anxiety and despair
(Leymann, 1990).

2002 The Scandinavian Psychological Associations. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA. ISSN 0036-5564.
398 E. G. Mikkelsen and S. Einarsen
According to transactional models of stress (e.g. Lazarus
& Folkman, 1984), the outcome of a stressful transaction is
mediated by the individuals appraisal and coping. Accord-
ingly, individual differences in victims appraisal and coping
with bullying may be one explanation as to why some
victims show a modest degree of impairment while others
exhibit severe symptoms of stress. Therefore, the present
study sought to investigate whether individual variables
pertaining to the victims appraisal and coping processes
may act as mediators and moderators of the relationships
between exposure to bullying behaviors at work and self-
reported psychological and psychosomatic health complaints.
Two factors that may act as such intervening variables are
state negative affect and generalized self-efficacy. These vari-
ables were the focus of this study.
It has been suggested that subjective emotional experience
or affective states are best depicted by two broad and
independent dimensions, namely positive affect and negative
affect (see George, 1995; Watson & Clark, 1984; Watson, Clark
& Tellegen, 1988). Positive affect indicates positive engage-
ment, reflecting the degree to which a person feels enthusi-
astic, active and alert (George, 1995; Watson & Clark, 1984).
Negative affect, on the other hand, is a general dimension of
subjective distress and is indicative of negative engagement,
reflecting the extent to which one is feeling upset or unpleas-
antly aroused ( Watson & Clark, 1984; Watson et al., 1988).
Whereas a state of low negative affect is captured by terms
such as calmness and serenity, individuals high in negative affect
are prone to feeling distressed, nervous, angry or guilty.
A persons degree of negative affect is a function of both
affective disposition (trait negative affectivity or trait-NA)
and life events (George, 1995). Trait-NA is a mood disposi-
tion that reflects pervasive individual differences in negative
emotionality and self-concept across time and different situ-
ations ( Watson & Clark, 1984). Individuals high in trait-NA
typically focus on the negative sides of life and tend to have
negative views of themselves, other people and the world in
general. Moreover, they are more likely to experience negat-
ive emotions such as distress, dissatisfaction, anxiety and
anger. Consequently, they will also be more likely to experi-
ence negative affect states. Indeed, a study by Brief, Burke,
George, Robinson and Webster (1988) showed that trait-NA
was significantly related to negative affect at work.
Exposure to negative life events is another significant con-
tributor to negative affect (George, 1995). Qualitative studies
( Kile, 1990; Mikkelsen, 1997) show that most victims of bul-
lying report negative feelings such as distress, nervousness,
anxiety and irritability due to being victimized at work.
Assuming that exposure to bullying is related to various
aversive mood states, a viable hypothesis is that such a state
of negative affect, or state negative affectivity (state-NA),
may mediate the relationship between exposure to bullying
behaviors and self-reported health complaints.
Following research within the field of victimology (e.g.
Janoff-Bulman, 1989), exposure to bullying behaviors at
2002 The Scandinavian Psychological Associations.
Scand J Psychol 43 (2002)
work may alter the victims emotional state and their views
of themselves and others, this in turn causing higher levels of
reported psychological health complaints and psychosomatic
complaints. Being a victim of intentional harm by another
person, real or perceived, seems to threaten peoples general
positive assumptions of themselves and the world and other
people (Janoff-Bulman, 1985), thus leading to severe emo-
tional reactions such as fear, anxiety, helplessness, depression
and shock. Victimization, in this case exposure to bullying
behaviors at work, may even change individuals perceptions
of their work environment to one of threat, danger, insecurity
and self-questioning (Janoff-Bulman & Frieze, 1983). An
elevation of state-NA might also initiate an increase in the
use of maladaptive coping strategies. These changes in the
perception of self and others as well as the accompanying
emotional state of mind may be long-lasting, especially if
accompanied by lack of social support at work or at home.
Part of the link between exposure to bullying and symp-
toms of stress may therefore be accounted for by an increase
in less favorable evaluations of self and others (Lakey, Tardiff
& Drew, 1994), again indicating an increase in state-NA.
It has also been suggested that a high degree of exposure to
prolonged job difficulties may lead to an increase in trait-
NA (Clark & Watson, 1991). Indeed, Lakey et al. (1994)
found that experiences of negative social interactions were
associated with an increase in trait-NA, while a lack of pos-
itive social interactions was not. Evidence that major stress-
ful life events may increase symptoms of stress by increasing
negative evaluations of others and self has been presented by
Lakey and Edmundson (1993) and may also be derived from
the work of Janoff-Bulman (1989).
Central to some definitions of workplace bullying is the
victims perception of being unable to cope with the negative
acts (Bjrkqvist et al., 1994; Einarsen, 2000; Keashly, 1998).
Accordingly, individuals low in self-esteem or self-efficacy
are likely to suffer more victimization than are others when
facing aggression or harassment (Einarsen, 2000). Low self-
esteem is found to be a consistent correlate of self-reported
exposure to bullying (Einarsen et al., 1996; Kile, 1990).
However, self-efficacy has not yet been studied in relation to
bullying at work.
The concept of self-efficacy refers to peoples beliefs about
their capacity to exercise control over events that affect their
lives (Bandura, 1992). Whereas self-efficacy is commonly
understood as referring to specific behaviors performed in
specific environmental contexts (Maddux, 1995), other
researchers (Jerusalem & Schwarzer, 1992) have introduced
a more trait-like version of the concept, termed generalized
self-efficacy. Generalized self-efficacy refers to stable and
global beliefs in the ability to deal efficiently with a wide
range of stressors (Jerusalem & Schwarzer, 1992; Schwarzer,
Bssler, Kwiatek, Schrder & Zhang, 1997). Generalized self-
efficacy may therefore be conceived as a personal resource
or vulnerability factor in the stress process, together with fac-
tors such as locus of control or self-esteem. In this respect
Scand J Psychol 43 (2002)
the construct is viewed as a moderator of the relationships
between stressors and individual outcomes.
Although studies have shown that self-efficacy may act as
a mediator of stressful experiences (see Litt, 1988), such a
mediator effect is less likely in relation to generalized self-
efficacy, which is a stable personality factor. Results from a
study of the role of victim personality in perceptions of sex-
ual harassment also pointed to self-efficacy as a moderator
inasmuch as self-efficacy did not influence how socio-sexual
behaviors in the workplace were perceived, but rather affected
the emotional reactions to instances of behavior commonly
seen as sexual harassment (Einarsen, Lillebrten & Roth,
1999). Women low on self-efficacy reported higher levels of
negative emotions when facing sexual harassment than
women high on self-efficacy.
The aim of the present study was therefore to investigate
the relationships between exposure to bullying behaviors at
work and psychological and psychosomatic health com-
plaints using a community sample of industrial workers. It
has been claimed that industrial workers are more at risk of
being exposed to workplace bullying (Einarsen & Raknes,
1997). With some exceptions (e.g. Bjrkqvist et al., 1994;
Einarsen & Raknes, 1997), many former studies in this field
have been based on self-selected victims of bullying. Since
self-selected victims may have been exposed to prolonged
and extreme degrees of bullying at work, these studies may
have demonstrated a somewhat exaggerated picture of the
relationships between bullying and health complaints. When
using a community sample of employees at an ordinary
workplace we will obtain information on the relationships
between health complaints and exposure to low-intensity
bullying behaviors or bullying in its first phases. On the
other hand, in such a sample we assumed that any mediating
effects of negative affect would be linked to state-NA as
opposed to trait-NA. A high level of victimization is prob-
ably required in order for a stable personality variable like
trait-NA to act as a mediator. Hence, the hypothesis of the
present study is that exposure to bullying behaviors may lead
to an increase in state-NA, which again may be related to an
increase in psychological and psychosomatic health com-
plaints. Moreover, since state-NA is not a stable personality
trait it cannot act as a moderator of the relationship
between exposure to bullying behaviors and strain reactions.
Accordingly, no such moderator hypothesis was studied.
Following the same line of argument, a mediating role of
generalized self-efficacy is theoretically implausible. In order
for such a mediator effect to take place a considerable degree
of exposure to bullying behaviors is required. Therefore, the
following hypotheses were tested in the present study:
H1. Exposure to bullying behaviors at work is associated
with an elevated level of psychological and psychosom-
atic health complaints.
H2. Exposure to bullying behaviors at work is associated
with increased levels of state-NA.
2002 The Scandinavian Psychological Associations.
Workplace bullying 399
H3. State-NA mediates the relationships between exposure
to bullying behaviors and psychological and psycho-
somatic health complaints.
H4. Generalized self-efficacy moderates the relationships
between bullying behaviors and psychological and psy-
chosomatic health complaints. The combination of a
high exposure to bullying behaviors and a low score on
generalized self-efficacy will show a significant relation-
ship with psychological and psychosomatic health com-
plaints over and above the relationships found between
these factors individually.
METHOD
Sample
Data were collected by means of anonymous questionnaires dis-
tributed to 433 employees in a Danish manufacturing company.
Such organizational settings have been found to constitute high-risk
situations for bullying (Einarsen & Skogstad, 1996). A total of
224 employees responded, a response rate of 51.7%. Although this
is a somewhat low response rate, Einarsen and Raknes (1997), who
succeeded in gathering information from respondents who initially
failed to complete the questionnaires, found that a low response rate
did not affect the results significantly. The sample comprised 55%
men and 45% women. Except for two persons, all employees were
between the ages of 19 to 59 years. Sixteen percent had only com-
pleted junior high school, 47% had finished their basic vocational
education or were educated as a trainee or apprentice. Ten percent
were skilled or specialized workers, while 26% had completed fur-
ther or higher education. Mean job tenure was 8.5 years (SD 7.13
years), ranging from 0 to 35 years. Thirty-eight of the respondents
(17%) were in a supervisory or managerial position.
Instruments
Exposure to bullying behaviors was measured by the Negative Acts
Questionnaire ( NAQ ) (Einarsen & Raknes, 1997). In this study the
scale comprises 18 items (Cronbachs alpha = 0.83) measuring
exposure to various negative acts. All items are described in behavioral
terms with no reference to the term bullying (see Table 1 for an
overview of the items included in this scale). This has the advantage
of measuring perceived exposure to behaviors without forcing the
respondents to label these behaviors as bullying. The method used
to chart exposure to bullying in the present study seems to prevent
an underestimation of the problem, which may occur when using
methods where individuals label themselves as victims of bullying
(Brooks & Perot, 1991). It also secures a somewhat more objective
description of the behaviors involved (see also Frese & Zapf, 1988).
The NAQ contains items referring to either direct (e.g. verbal
abuse, offensive remarks, ridicule) or indirect (e.g. social ostracism,
slander) behaviors. For each item of the NAQ the respondents were
asked how often they had been exposed to the behavior during the
last six months, response categories being: never, now or then, about
weekly, about daily. A weekly exposure to such negative acts over a
period of six months has been proposed as an operational definition
of victimization due to bullying at work (Leymann, 1996).
Psychological health complaints were measured by the Hopkins
Symptom Checklist (HSCL-25), originally developed by Derogatis,
Lipman, Rickels, Uhlenhuth and Covi (1974). The scale measures
psychiatric symptoms, including symptoms of anxiety, depression
and somatization. The respondents were asked to indicate to what
400 E. G. Mikkelsen and S. Einarsen
extent they had felt distressed or bothered by each symptom within
the previous week. All scale items are scored on a five-point scale:
very slightly or not at all, a little bit, moderately, quite a bit or
extremely. The HSCL-25 has excellent internal consistency, with a
Cronbachs alpha of 0.92. In the present study it was used as a
measurement of psychological health complaints.
Psychosomatic complaints were measured by a seven-item psy-
chosomatic scale taken from the Stress Profile developed by Setter-
lind and Larsson (1995). The respondents were asked to report the
perceived frequency of various psychosomatic symptoms within
the last four weeks, such as dizziness, stomachache and chest pain.
The items are scored on a five-point scale ranging from always to
never. In the present study the internal consistency of the scale was
satisfactory (Cronbachs alpha = 0.85).
State-NA was measured by the 10-item NA scale of the PANAS
Scales originally developed by Watson, et al. (1988). The respond-
ents were asked to indicate to what extent they had felt a particular
feeling or emotion within the last two weeks, such as being dis-
tressed, afraid, guilty, or nervous. Response categories were very
slightly, a little, moderately, quite a bit, and extremely. In the pre-
sent study, the scale had an acceptable internal consistency, with a
Cronbachs alpha of 0.79.
The 10-item General Perceived Self-Efficacy Scale (Jerusalem &
Schwarzer, 1992) measured generalized self-efficacy. The scale has
been shown to have acceptable reliability and evidence of validity
has been provided. The following are examples of the items on this
scale: I can always manage to solve difficult problems if I try hard
enough; I am confident that I could deal efficiently with un-
expected events. For each statement the respondents were asked to
indicate the extent to which they considered it characterised them
as persons. The response categories were not at all true, barely true,
moderately true or exactly true. In this particular study, the scale
had an excellent internal consistency (Cronbachs alpha = 0.89).
Translations of the HSCL-25, the PANAS Scales and the General
Perceived Self-Efficacy Scale (from the English versions) were done
in accordance with the requirements adhering to scientific work:
translation and back-translation by bilingual consultants with a
university degree in psychology and / or English.
Statistical procedures
Pearsons productmoment correlations and partial correlations
were computed in order to measure the degree of associations
between exposure to bullying behaviors at work, state-NA, gen -
eralized self-efficacy and the health measures. In order to test for
possible moderating effects of generalized self-efficacy we used
hierarchical regression analyses following the prescriptions of Baron
and Kenny (1986). In all the statistical analyses, exposure to bully-
ing behaviors was measured by a sum-scale of the NAQ ( Einarsen
& Raknes, 1997).
To ensure discriminatory validity for the measurements of psy-
chosomatic complaints, psychological health complaints and state-
NA, a series of exploratory factor analyses were conducted using
principal-component analysis with varimax rotation. First of all, a
principal-component analysis with varimax rotation showed the
three scales involved to be distinct and separate factors. Secondly,
a separate analysis for state-NA and psychosomatic complaints
showed that seven of the state-NA items loaded on their own factor
with no problematic cross-loadings. That is, they loaded above 0.60
on their own factor, while loading lower than 0.35 on the corres-
ponding psychosomatic factor. On the latter, all items loaded
exclusively on their own factor. The same seven state-NA items
showed an identical relationship when being factor analyzed
together with HCSL-25. Hence, this seven-item version of state-NA
has been used in the analyses of the present study, ensuring the
2002 The Scandinavian Psychological Associations.
Scand J Psychol 43 (2002)
investigation of potentially related but distinct concepts. The follow-
ing items were included in this revised version of state-NA: afraid,
guilty, nervous, distressed, ashamed, scared and jittery. The
excluded items were: upset, hostile, and irritable. This revised ver-
sion of the state-NA scale had a high internal consistency, with a
Cronbachs alpha of 0.87.
RESULTS
A total of 88% of the respondents reported exposure to at
least one of the behaviors measured by the NAQ during the
previous six months. However, 8% reported exposure to at
least one bullying act weekly or more often during the pre-
vious six months, thus being potential victims of bullying
according to the criteria suggested by Leymann (1996).
Exposure to two different types of behaviors weekly or more
often were reported by 2.7% of the respondents. Table 1
shows the incidence of the various bullying acts in this
sample. Chi-square tests showed no significant gender differ-
ences in relation to exposure to bullying acts (2 = 22.29,
d.f. = 19, p = 0.267).
Pearsons productmoment correlations were computed
between exposure to bullying behaviors, psychological
health complaints, psychosomatic complaints, state-NA
and generalized self-efficacy (Table 2). Exposure to bullying
behaviors at work correlated strongly with psychological
health complaints (r = 0.52, p < 0.01) and moderately with
psychosomatic complaints (r = 0.32, p < 0.01). The direc-
tions of the correlations showed that reported exposure to
bullying behaviors was associated with more psychological
health complaints (i.e. a lower level of psychological well-
being) and more psychosomatic complaints, although the
latter association was more moderate. These results thus give
support to hypothesis 1. Also, in agreement with hypo-
thesis 2, exposure to bullying behaviors correlated strongly
with state-NA. Generalized self-efficacy, however, was not
associated with exposure to bullying behaviors. State-NA
also correlated rather strongly with both psychological and
psychosomatic health complaints.
Using partial correlation, results were consistent with
state-NA as a mediator of relationships between exposure to
bullying behaviors and both psychological health complaints
and psychosomatic complaints. Yet, state-NA acted only as
a partial mediator of the relationships. When controlling for
the effects of state-NA, the predictorcriterion relationship
between bullying behaviors and psychological health com-
plaints was reduced, but remained statistically significant
(r = 0.52, pr = 0.30, p < 0.01). Having controlled for state-
NA, the relationship between bullying behaviors and psy-
chosomatic complaints was also reduced, but still significant
(r = 0.32, pr = 0.19, p < 0.05). Demographic variables such
as gender, age and educational level did not influence
these relationships. Thus, these results are supportive of
hypothesis 3.
Although not related to exposure to bullying behaviors in
itself, generalized self-efficacy seemed to act as a moderator
Scand J Psychol 43 (2002) Workplace bullying 401

Table 1. Percentage of employees endorsing each item on the NAQ

During the last six months, how often have you been Now About About
subjected to the following negative acts in the work place? Never and then weekly daily

1. Someone withholding necessary information affecting your performance 36.7% 57.5% 5.0% 0.9%
2. Insulting teasing 82.4% 16.7% 0.9%
3. Being ordered to do work below your level of competence 55.0% 37.3% 4.5% 3.2%
4. Being deprived of responsibility or work tasks 77.8% 21.3% 0.9%
5. Slander or rumors about you 67.0% 31.2% 1.4% 0.5%
6. Social exclusion from co-workers or work group activities 80.6% 18.5% 0.9%
7. Repeated offensive remarks about your person or your private life 86.9% 11.3% 0.9% 0.9%
8. Verbal abuse 80.6% 18.5% 0.5% 0.5%
9. Ridicule 85.1% 14.0% 0.9%
10. Hints or signals from others that you should quit your job 97.3% 1.8% 0.9%
11. Repeated reminders of your blunders 77.0% 22.1% 0.9%
12. Silence or hostility as a response to your questions or attempts at conversations 84.6% 14.5% 0.9%
13. Devaluation of your work and efforts 80.6% 18.0% 0.9% 0.5%
14. Neglect of your opinions or views 61.7% 35.6% 2.7%
15. Devaluation of your rights and opinions with reference to your gender 92.3% 7.2% 0.5%
16. Devaluation of your rights and opinions with reference to your age 94.1% 5.4% 0.5%
17. Exploitation at work, such as private errands 95.0% 4.5% 0.5%
18. Negative reactions from others because you work too hard 68.3% 27.6% 4.1%

Table 2. Intercorrelations between psychosomatic complaints, HCSL- and the cross-product term were entered together, the latter
25 score, bullying, generalized self-efficacy and state-NA accounted for an extra two percentage points of the vari-
ance, which was a significant increase (p < 0.05) (R2 = 0.33,
Scale 1 2 3 4 5
F = 6.25, p < 0.05, for the total model). Even though this
1. Bullying 0.41** 0.07 0.32** 0.52** effect is indeed very moderate, the results do indicate that
2. State-NA revised 0.27** 0.41** 0.70** individuals with a higher degree of generalized self-efficacy
3. Generalized 0.20** 0.24** will report fewer psychological health complaints than low
self-efficacy
self-efficacious individuals following exposure to bullying
4. Psychosomatic 0.65**
complaints behaviors.
5. HSCL-25 At the same time, generalized self-efficacy did not moder-
ate the relationship between exposure to bullying behaviors
**p < 0.01. and psychosomatic complaints. A multiple regression ana-
lysis showed that only exposure to bullying behaviors
Table 3. Hierarchical regression analysis testing the moderating effect accounted for a partial explained variation in psychosomatic
of generalized self-efficacy on the relationship between bullying and complaints (R2 = 0.10, p < 0.05). Neither self-efficacy nor
psychological health complaints the interaction term added any extra explained variance. Put
together, these results give partial support to hypothesis 4,
R2 F- total F-change.
in as much as generalized self-efficacy acted as a moderator
Bullying 0.27 78.74*** of the relationship between exposure to bullying behaviors
+ Generalized self-efficacy 0.31 48.29*** 12.78** and psychological health complaints but not psychosomatic
+ Generalized self-efficacy bullying 0.33 33.83*** 6.57* complaints. Demographic variables such as gender, age and
educational level did not influence any of these relationships
*p < 0.05, **p < 0.01, ***p < 0.001.
when they were entered as control variables in the multiple
regression analyses.

of the relationship between reported bullying behaviors and

psychological health complaints ( Table 3). A hierarchical
multiple regression analysis showed that exposure to bully-
ing behaviors explained 27% of the variance in psychological
health complaints, whereas the introduction of generalized
self-efficacy in the regression model added a significant
increase of four percentage points in the explained variance
(R2 = 0.31, p < 0.01). However, when both individual variables
DISCUSSION
The fact that 88% of the respondents reported being
exposed to aggressive interpersonal acts now and then dur-
ing the last six months indicates that such acts are relatively
common in this organizational setting. Identical results were
found in a comparable Norwegian study (Einarsen &
Raknes, 1997). Inasfar as such acts occur regularly, they may

2002 The Scandinavian Psychological Associations.

402 E. G. Mikkelsen and S. Einarsen
be indicative of bullying. The relatively high endorsement of
items involving personal disparagement for example expos-
ure to slander or rumors, neglect of ones opinion or points
of views, exposure to ridicule and insulting teasing as well
as other kinds of offensive remarks, is a potential indicator
of the severity of the problem. Several studies (Einarsen &
Raknes, 1997; Zapf et al., 1996) have shown that behaviors
experienced as personal disparagement are particularly
offensive and therefore strongly associated with lowered
wellbeing and subjective health complaints.
The results of the study offered empirical support for our
first hypothesis. Exposure to bullying behaviors at work is
significantly associated with increased levels of both psycho-
logical health complaints and psychosomatic complaints.
Whereas reported exposure to bullying behaviors alone
accounted for 27% of the variance in psychological health
complaints, the 10% explained variance in psychosomatic
complaints was more moderate. Although the design of the
study does not provide a test of the causal ordering of the
constructs, the study provides further empirical support for
the assertion that bullying is an important work environ-
ment stressor, with strong associations with poor victim
health and wellbeing (Bjrkqvist et al., 1994; Brodsky, 1976;
Einarsen & Raknes, 1997; Niedl, 1996; Zapf et al., 1996).
The fact that the perceived behaviors involved in bullying on
the one hand seem to occur fairly often, and on the other
hand are highly associated with health complaints when
occurring regularly, is also consistent with the work of
Leymann (1990). However, the cross-sectional nature of the
data limits causal conclusions on the relationships between
exposure to bullying behaviors and health.
In support of our second hypothesis, state-NA was found
to correlate strongly with exposure to bullying acts. Fur-
thermore, the results indicated that state-NA might act as
a partial mediator of the relationships between exposure
to bullying behaviors and both psychological and psycho-
somatic health complaints, thus partially corroborating
hypothesis 3. Given the possibility that state-NA acts as a
partial mediator, thus accounting for a portion of the rela-
tion between the predictor variable and the criterion variable
by explaining how external physical events take on a psycho-
logical meaning (Baron & Kenny, 1986; Cox & Ferguson,
1993), exposure to bullying behaviors may result in
enhanced levels of state-NA. This might initiate an increased
use of maladaptive coping strategies, in turn causing higher
levels of reported psychological health complaints and psy-
chosomatic complaints.
Exposure to bullying behaviors and state-NA may also
interact in a vicious circle of events. Exposure to interper-
sonal problems and conflicts may of course bring about an
elevated level of distress in the individual. Distressed indi-
viduals may then interpret the behavior of others as personal
insults or attacks, thus increasing their own level of negative
emotions, which again may result in others avoiding them or
reacting aggressively to them. Exposure to bullying beha-
2002 The Scandinavian Psychological Associations.
Scand J Psychol 43 (2002)
viors may both justify and enhance their negative emotional
state as well as their negative attitude to others.
However, due to the correlational design of the study,
alternative explanations must also be forwarded. First of all,
it may be the case that a more stable trait-NA is a cause of
both the observed state-NA as well as the perceptions of
bullying behaviors and health complaints. Trait-NA has
proved to be strongly related to state-NA even when the
state measures are completed 10 years after the trait meas-
ures (Watson & Clark, 1984). Burke, Brief and George
(1993, p. 405) also point to the trait-like stability of state-
NA, concluding that when respondents are instructed to
rate their negative affect over the past few weeks, such ratings
can possibly be used as a measure of negative affectivity.
Spector, Zapf, Chen and Frese (2000) have outlined six
mechanisms by which trait-NA may play a substantive role
in relation to job stressors and job strains: the perception
mechanism, the hyper-responsivity mechanism, the selec-
tion mechanism, the stressor creation mechanism, the mood
mechanism and the causality mechanism. We suggest that at
least three of these mechanisms may help explain why trait-
NA might affect the relationship between exposure to bully-
ing behaviors at work and reported strain reactions.
Firstly, consistent with the perception mechanism suggested
by Spector et al. (2000), trait-NA may affect employees
perceptions of the social climate and consequently the
way they perceive interpersonal behaviors. High trait-NA
individuals tend to focus on the negative sides of themselves,
other people and the world. Also, they are more inclined to
give attention to and magnify potential threats from their
surroundings and to perceive ambiguous stimuli negatively
(Watson & Clark, 1984). Accordingly, they may be more
likely to perceive the behavior of others as acts of bullying.
This perceived exposure to bullying behaviors might in turn
be related to increased levels of reported strain.
Secondly, in agreement with the hyper-responsivity
hypothesis (Spector et al., 2000), individuals high on trait-
NA may have an increased emotional reactivity when facing
stressors such as negative interpersonal behaviors (Larsen &
Ketelaar, 1991). Hence, when exposed to bullying behaviors
they may experience more strain than others given the same
environmental conditions (Spector et al., 2000).
Finally, according to the stressor creation mechanism
(Spector et al., 2000), high trait-NA individuals may in fact
create a social environment in which they are exposed to
aggressive behaviors. For instance, the behavior of high trait-
NA individuals may annoy and irritate others, who may
then react negatively toward them (Burke et al., 1993; Zapf,
1999a). From a theoretical point of view, the particular con-
stellation of the high trait-NA individual, that is, their
proneness to experiencing feelings of distress, anxiety and
anger at any time and across a variety of situations, together
with their introvert, non-conformist and slightly anti-social
personality, could make them targets for bullying (Felson &
Tedeschi, 1993).
Scand J Psychol 43 (2002)
Moreover, the tendency toward the use of coping strateg-
ies involving confrontation, for example by angrily venting
emotions as a response to stress (OBrien & DeLongis,
1996), might also result in high trait-NA individuals being
more frequently involved in interpersonal conflicts, which
again may escalate into bullying (see Einarsen, 2000). This
line of argument may be in accordance with findings of stud-
ies of bullying in schools in which researchers have described
certain types of victims as provocative and annoying
(Olweus, 1987). In this respect it is worth noting that,
although distressed persons may themselves alter their own
work environment in a negative direction, their perceptions
of this altered work environment may still be valid.
It may also be argued that a conceptual kinship exists
between the concept of state-NA and psychological health
complaints as measured by the HSCL-25. However, when
controlling for the effect of state-NA, the relationship
between exposure to bullying behaviors and psychological
health complaints remained statistically significant. Factor
analysis also showed these concepts to be distinct. In our
opinion, this result suggests that exposure to bullying beha-
viors represents a significant job stressor with relationships
to psychological health complaints over and above what may
be explained by a dispositional or, alternatively, a changed
level of state-NA within the victim.
Contrary to state-NA, generalized self-efficacy showed no
association with exposure to bullying behaviors in the
present study. Intuitively, this may appear somewhat surpris-
ing as one would expect people with a low sense of general-
ized self-efficacy to be more exposed to aggressive behaviors
from others. For example, studies have shown that people
who lack self-confidence experience more stress and perform
less competently as a response to stressors ( Bandura, 1992;
Jerusalem & Schwarzer, 1992). From a theoretical point of
view, this may increase the risk of being subjected to aggres-
sion (Felson & Tedeschi, 1993). Prior studies have also
pinpointed the negative effects of bullying on the victims
sense of self-competence ( Leymann, 1990). However, it may
be questioned whether the General Perceived Self-Efficacy
Scale used in this study sufficiently captures aspects of social
competence essential to the management of interpersonal
conflicts or aggression. Only one item refers explicitly to social
situations. A domain-specific scale for self-efficacy may have
generated different results (see also Schwarzer et al., 1997).
Conversely, several Scandinavian researchers (e.g. Einarsen,
2000; Thylefors, 1987 ) have argued that many victims of
bullying are in fact competent and resourceful persons who
just happen to be involved in a harsh work-related conflict.
According to a German study (Zapf, 1999a), such an expla-
nation seems to be correct for a large number of victims.
At the same time, generalized self-efficacy did moderate
the relationship between exposure to bullying behaviors
and psychological health complaints, although the effect was
small. When being exposed to bullying behaviors, it is likely
that highly self-efficacious individuals will feel more capable
2002 The Scandinavian Psychological Associations.
Workplace bullying 403
of handling the situation, for example by rebuking the
aggressor or by seeking support from colleagues or super-
iors. In contrast, low self-efficacious individuals may be more
prone to anticipating a failure scenario (Bandura, 1992).
Consequently, they may feel incapable of dealing with the
bullying behaviors and may be more likely to use coping
strategies typically conceived of as more maladaptive. In
fact, several studies have shown low self-efficacy beliefs to be
associated with an increased use of emotion-focused cop-
ing (Chwalisz, Altmaier & Russell, 1992; Stumpf, Brief &
Hartman, 1987), which in turn have been related to greater
symptoms of stress (Lazarus & Folkman, 1984). The felt
inability to cope with negative social interactions may in itself
create a strong feeling of being victimized when exposed to
aggression. Several researchers consider the victims per-
ceived inability to cope as a key element in the process of
victimization of bullying (Einarsen, 2000; Niedl, 1995; Zapf,
1999a). The felt inability to cope is what leads a potentially
stressful circumstance to be perceived as a real threat
(Lazarus & Folkman, 1984).
Interestingly, we found only a weak moderator effect of
generalized self-efficacy. One explanation for this finding
might be that, in many cases, what the victims actually do
in order to hinder further exposure to bullying might be
extraneous to the outcome. Up to now, only a few studies
(Keashly et al., 1994; Niedl, 1996; Rayner, 1998; Thylefors,
1987) have focused on victims ways of coping with bullying.
These studies show that the victims use a variety of strat-
egies, including both problem-orientated strategies (talking
with or confronting the aggressor, collecting evidence,
seeking support from superiors or colleagues etc.) and
emotion-focused or avoidance strategies (distancing, denial, self-
control, self-abasement etc.). However, even though some
victims report that problem-focused strategies may alleviate
the situation considerably (Keashly et al., 1994), in many
instances they are to no avail (Niedl, 1996; Rayner, 1998)
or may even lead to deterioration of the situation (Zapf,
1999b). Consequently, because exposure to bullying, especi-
ally long term, typically causes the victim to feel helpless
and out of control, and since these subjective feelings often
mirror the reality of a bullying process, personality factors
may be of less importance in determining coping strategies
and hence health outcomes. Nonetheless, another and pre-
sumably more feasible explanation for the weak moderator
effect of generalized self-efficacy in this study must also be
posited: Given that few of the employees were subjected to
severe bullying, the effects of individual differences in per-
ceived ability to cope may be moderate.
CONCLUSION
Results from the present study confirm previous findings
showing that exposure to bullying behaviors is associated
with an increase in psychological health complaints and
increased levels of psychosomatic health complaints. While
404 E. G. Mikkelsen and S. Einarsen
the present paper explored health correlates of exposure to
bullying behaviors, future studies should look more closely
into the likely differences in reported strain reactions among
those who label themselves as victims of bullying and those
who only report exposure to bullying acts.
Furthermore, results of the study were indicative of state-
NA as a partial mediator of the relationships between expos-
ure to bullying behaviors and self-reported strain reactions.
Generalized self-efficacy, on the other hand, seemed to
moderate the relationship between exposure to bullying
behaviors and psychological health complaints. These
results point to the relevance of investigating individual
mediator and moderator variables in relation to explaining
differences in reported strain following exposure to bullying
at work. The present study focused only on the effects of
what we believe to be two potentially important variables,
namely state-NA and generalized self-efficacy. Further studies
investigating the effects of these and other variables pertain-
ing to individual differences are recommended.
REFERENCES
Bandura, A. (1992). Exercise of personal agency through the self-
efficacy mechanism. In R. Schwarzer (Ed.), Self-efficacy: Thought
control of action. New York: Hemisphere.
Baron, R. M. & Kenny, D. A. (1986). The moderatormediator
variable distinction in social psychological research: Conceptual,
strategic, and statistical considerations. Journal of Personality
and Social Psychology, 51, 11731182.
Bjrkqvist, K., sterman, K. & Hjelt-Bck, M. (1994). Aggression
among university employees. Aggressive Behavior, 20, 173 184.
Brief, A. P., Burke, M. J., George, J. M., Robinson, B. S. & Webster, J.
(1988). Should negative affectivity remain an unmeasured vari-
able in the study of job stress? Journal of Applied Psychology,
73, 193198.
Brodsky, C. (1976). The harassed worker. Toronto: Lexington
Books, DC Heath and Company.
Brooks, L. & Perot, A. R. (1991). Reporting sexual harassment:
Exploring a predictive model. Psychology of Women Quarterly,
15, 31 47.
Burke, M. J., Brief, A. P. & George, J. M. (1993). The role of negat-
ive affectivity in understanding relations between self-reports of
stressors and strains: A comment on the applied psychology lit-
erature. Journal of Applied Psychology, 78, 402 412.
Chen, P. Y. & Spector, P. E. (1991). Negative affectivity as the
underlying cause of correlations between stressors and strains.
Journal of Applied Psychology, 76, 338 407.
Chwalisz, K., Altmaier, E. M. & Russell, D. W. (1992). Causal attri-
butions, self-efficacy cognitions and coping with stress. Journal
of Social and Clinical Psychology, 11, 377 400.
Clark, L. A. & Watson, D. (1991). General affective dispositions in
physical and psychological health. In C. R. Snyder & D. R.
Forsyth (Eds.), Handbook of social and clinical psychology:
The health perspective (pp. 221 245). New York: Pergamon Press.
Cox, T. & Ferguson, E. (1993). Individual differences, stress and
coping. In C. L. Cooper & R. Payne ( Eds.), Personality and
stress: Individual differences in the stress process. Chichester:
John Wiley & Sons.
Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H. &
Covi, L. (1974). The Hopkins Symptom Checklist (HSCL): A
self-report symptom inventory. Behavioral Science, 19, 1 15.
2002 The Scandinavian Psychological Associations.
Scand J Psychol 43 (2002)
Einarsen, S. (2000). Harassment and bullying at work: A review of
the Scandinavian approach. Aggression and Violent Behavior: a
Review Journal, 4, 371 401.
Einarsen, S., Lillebrten, G. & Roth, L. (1999). The role of person-
ality in perceptions of sexual harassment. Paper submitted for
publication.
Einarsen, S., Matthiesen, S. B. & Skogstad, A. (1998). Bullying,
burnout and well-being among assistant nurses. Journal of Occu-
pational Health and Safety Australia and New Zealand, 14,
563568.
Einarsen, S. & Raknes, B. I. (1997 ). Harassment in the work-
place and the victimization of men. Violence and Victims, 12,
247 263.
Einarsen, S., Raknes, B. I., Matthiesen, S. B. & Hellesy, O. H.
(1996). Helsemessige aspekter ved mobbing i arbeidslivet. Mod-
ererende effekter av social sttte og personlighet. (Bullying at
work and its relationships with health complaints. Moderating
effects of social support and personality.) Nordisk Psykologi, 48,
116 137.
Einarsen, S. & Skogstad, A. (1996). Bullying at work: Epidemiolo-
gical findings in public and private organizations. European
Journal of Work and Organizational Psychology, 5, 185201.
Felson, R. B. & Tedeschi, J. T. (1993). Aggression and Violence
Social interactionist Perspectives. Washington, DC: American
Psychological Association.
Frese, M. & Zapf, D. (1988). Methodological issues in the study of
work stress: Objective vs. Subjective Measurements and the
question of longitudinal Studies. In C. L. Cooper and R. Payne
(Eds.), Causes, coping and consequences of stress at work
( pp. 371 411). Chichester: Wiley and Sons.
George, J. M. (1995). Trait and state affect. In K. R. Murphy (Ed.),
Individual differences and behavior in organizations (pp. 145
171). San Francisco: Jossey-Bass.
Hoel, H., Rayner, C. & Cooper, C. L. (1999). Workplace bullying.
International Review of Industrial and Organizational Psychology,
14: 195 230.
Janoff-Bulman, R. (1985). Criminal vs. non-criminal victimization:
Victims reactions. Victimology: An International Journal, 10,
498511.
Janoff-Bulman, R. (1989). Assumptive worlds and the stress of
traumatic events: Applications of the schema construct. Social
Cognition, 7, 113136.
Janoff-Bulman, R. & Frieze, I. H. (1983). A theoretical perspective
for understanding reactions to victimization. Journal of Social
Issues, 39, 117.
Jerusalem, M. & Schwarzer, R. (1992). Self-efficacy as a resource
factor in the stress appraisal processes. In R. Schwarzer (Ed.),
Self-efficacy: Thought control of action. New York: Hemisphere.
Keashly, L. (1998). Emotional abuse in the workplace: Conceptual
and empirical issues. Journal of Emotional Abuse, 1, 85117.
Keashly, L., Trott, V. & MacLean, L. M. (1994). Abusive behavior
in the workplace: a preliminary investigation. Violence and Vic-
tims, 9, 341357.
Keenan, A. & Newton, T. J. (1984). Frustrations in organizations:
Relationships to role stress, climate and psychological pain.
Journal of Occupational Psychology, 57, 57 65.
Kile, S. M. (1990). Helsefarleg leiarskap. Ein eksplorerande studie.
Rapport til Norges Almenvitskaplege forskningsrd. Universite-
tet i Bergen: Institutt for Samfunnspsykologi.
Lakey, B. & Edmundson, D. D. (1993). Role evaluations and stress-
ful life events: Aggregate versus domain-specific predictors. Cog-
nitive Therapy and Research, 17, 249267.
Lakey, B., Tardiff, T. A. & Drew, J. B. (1994). Negative social inter-
actions: Assessment and relations to social support, cognition,
and psychological distress. Journal of Social and Clinical Psy-
chology, 13, 42 62.
Scand J Psychol 43 (2002)
Larsen, R. J. & Ketelaar, T. (1991). Personality and susceptibility to
positive and negative emotional states. Journal of Personality
and Social Psychology, 61, 132140.
Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal and coping.
New York: Springer Publishing Company.
Leymann, H. (1990). Mobbing and psychological terror at work-
places. Violence and Victims, 5, 119126.
Leymann, H. (1996). The content and development of mobbing at work.
European Journal of Work and Organizational Psychology, 5, 165 184.
Litt, M. D. (1988). Cognitive mediators of stressful experience:
Perceived self-efficacy and perceived control. Cognitive Therapy
and Reseach, 12, 241260.
Maddux, J. E. (1995). Self-efficacy theory: An introduction. In
J. E. Maddux (Ed.), Self-efficacy, adaptation and adjustment.
New York: Plenum Press.
Mikkelsen, E. G. (1997). Mobning i arbejdslivet. En krise og arbejd-
spsykologisk undersgelse. ( Workplace bullying. An analysis of
its trauma related and occupational psychological aspects.)
Unpublished masters Thesis, Department of Psychology, Uni-
versity of Aarhus.
Niedl, K. (1995). Mobbing / bullying am arbeitsplatz. Mnchen:
Rainer. Hampp Verlag.
Niedl, K. (1996). Mobbing and well-being: Economic and personnel
development implications. European Journal of Work and Organ-
izational Psychology, 5, 239249.
OBrien, T. B. & DeLongis, A. (1996). The interactional context of
problem-, emotion-, and relationship-focused coping: The role of the
Big Five personality factors. Journal of Personality, 64, 774 813.
Olweus, D. (1987). Bully/victim problems among schoolchildren
in Scandinavia. In J. P. Myklebust and R. Ommundsen
(Eds.), Psykologprofesjonen mot r 2000 ( pp. 395 413). Oslo:
Universitetsforlaget.
Rayner, C. (1998). Workplace bullying: Do something! Journal of
Occupational Health and Safety Australia and New Zealand,
14, 581585.
2002 The Scandinavian Psychological Associations.
Workplace bullying 405
Schwarzer, R., Bssler, J., Kwiatek, P., Schrder, K. & Zhang, J. X.
(1997). The assessment of optimistic self-beliefs: Comparison of
the German, Spanish and Chinese versions of the General Self-
Efficacy Scale. Applied Psychology: An International Review, 46,
6988.
Setterlind, S. & Larsson, G. (1995). The Stress Profile: a psycho-
social approach to measuring stress. Stress Medicine, 11, 8592.
Spector, P. E., Zapf. D., Chen, P. Y. & Frese, M. (2000). Why neg-
ative affectivity should not be controlled in job stress research:
Dont throw the baby out with the bath water. Journal of Organ-
izational Behaviour, 21, 7995.
Stumpf, S. A., Brief, A. P. & Hartman, K. (1987). Self-efficacy
expectations and coping with career-related events. Journal of
Vocational Behavior, 31, 91108.
Thylefors, I. (1987). Syndabockar Om utsttning och mobbning i
arbetslivet. Stockholm: Natur och Kultur.
Watson, D. & Clark, L. A. (1984). Negative affectivity: The dis-
position to experience aversive emotional states. Psychological
Bulletin, 96, 465 490.
Watson, D., Clark, L. A. & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The
PANAS Scales. Journal of Personality and Social Psychology, 54,
1063 1070.
Zapf, D. (1999a). Organizational, work group related and personal
causes of mobbing/bullying at work. International Journal of
Manpower, 1, 70 85.
Zapf, D. (1999b). Mobbing in organisationen. Ein uberblick zum
stand der forschung. Zeitschrift fur Arbeids- & Organisation-
spsychologie, 43, 85 100.
Zapf, D., Knorz, C. & Kulla, M. (1996). On the relationship
between mobbing factors, and job content, social work environ-
ment and health outcomes. European Journal of Work and
Organizational Psychology, 5, 215 237.
Received 24 January 2001, accepted 16 July 2001