You are on page 1of 19

Work & Stress

Vol. 26, No. 3, JulySeptember 2012, 195212

The impact of role stressors on workplace bullying in both victims


and perpetrators, controlling for personal vulnerability factors:
A longitudinal analysis
Cristian Balduccia*, Monica Cecchinb and Franco Fraccarolic
a

Department of Political and Social Sciences, University of Bologna, Italy; bULSS Asolo, Asolo,
Italy; cDepartment of Cognitive and Education Sciences, University of Trento, Italy

Although it has been suggested that a poor work environment can be related to the incidence
of bullying, little work with robust research designs has been conducted on the matter. By
drawing on the concept of hindrance stressors and using a longitudinal research design, we
investigated whether role conflict and role ambiguity predicted being a victim of bullying
twelve months later, over and above personal vulnerability factors. With a parallel analysis we
also investigated whether the same role stressors predicted the enactment of bullying. The
sample consisted of 234 employees of a National Health Service agency in Italy, including
medical, nursing and administrative staff. The results indicated that role conflict positively
affected both being bullied and bullying enactment, with personal vulnerability (reporting a
doctors diagnosis of depression at baseline) affecting only the first of the two outcomes.
However, some evidence also emerged of reciprocal relationships between role stressors and
bullying. Directions for future research on the relations between working conditions and
bullying are discussed.
Keywords: workplace bullying; harassment; aggressive behaviour; role stressors; depression;
personality

Introduction
Workplace bullying is an extreme psychosocial risk at work in which a number of
negative behaviours, such as the withholding of information that affects performance, the spreading of rumours, social isolation and verbal abuse are frequently
and persistently directed over time at one individual employee (Einarsen, Hoel, Zapf,
& Cooper, 2011; Leymann, 1996). In the longer run, these behaviours may lead to the
stigmatization and victimization of the exposed individual, resulting in extreme stress
reactions (e.g. Mikkelsen & Einarsen, 2002). Concepts such as emotional abuse,
victimization and harassment (for a review see Aquino & Thau, 2009), mainly
adopted by North American researchers, have also been used in this area of research
to study phenomena that share most of their defining characteristics with bullying,
the main difference being that the concept of bullying emphasizes the repetitive

*Corresponding author. Email: cristian.balducci3@unibo.it


ISSN 0267-8373 print/ISSN 1464-5335 online
# 2012 Taylor & Francis
http://dx.doi.org/10.1080/02678373.2012.714543
http://www.tandfonline.com

196

C. Balducci et al.

nature (i.e. on a weekly or daily basis) and prolonged duration (i.e. several months) of
the negative workplace behaviours (Einarsen et al., 2011).
Most of the early research on bullying was devoted to refining the construct and
understanding its consequences. There is now substantial agreement, at least among
European researchers, on how to define and operationalize bullying (see above). As
far as its consequences are concerned, being a victim of bullying has been related to a
number of negative health conditions such as clinically-relevant anxiety and
depression (see Hogh, Mikkelsen, & Hansen, 2011). A positive and significant
relationship has also been reported between exposure to bullying and suicidal
ideation and behaviour (Balducci, Alfano, & Fraccaroli, 2009). These results have led
some researchers (Zapf & Einarsen, 2005) to argue that the consequences of bullying
are more extreme than those usually reported for other psychosocial risks in the
workplace.
More recent research on bullying has adopted a preventive stance, and it has
begun to focus on its potential causes. One hypothesis that has received increasing
attention in this regard emphasizes the role of a poor work environment as a critical
antecedent of bullying (Einarsen, Raknes, & Matthiesen, 1994; Leymann, 1996).
Various processes may explain how the work environment may promote bullying. In
line with a social interactionist perspective (Neumann & Baron, 2011), it has been
proposed that poor working conditions may lead future victims, through ineffective
coping, to behavioural reactions such as violations of social norms (e.g. treating
others in a disrespectful manner, constantly complaining about the situation,
withdrawal  see also Baillien, Neyens, De Witte, & de Cuyper, 2009), which may
induce coercive reactions against them in the form of bullying. Alternatively, the
negative work environment reported by victims may be indicative of the presence of
ambient stressors that are perceived also by the perpetrators, who enact bullying in
response to those stressors (Bowling & Beehr, 2006). A further explanation, which is
similar to the latter, but which focuses exclusively on bullying enactment, postulates
that bullying behaviour is a reaction to emotionally-critical internal states (e.g. anger,
anxiety) generated by poor working conditions. This explanation builds upon the
frustration-aggression hypothesis as reformulated by Berkowitz (1989), which has
been adopted in this area (see Spector & Fox, 2005) to investigate interpersonal
counterproductive work behaviour, a specific form of deviant behaviour that is very
similar to bullying (e.g. Balducci, Schaufeli, & Fraccaroli, 2011). The common
denominator of these various explanations is that bullying is a strain phenomenon
triggered by poor working conditions.
Research based on the work environment hypothesis has increased in the past
decade. In most cases, studies have focused on the experience of being bullied.
Skogstad, Einarsen, Torsheim, Aasland, and Hetland (2007), for example, found that
role conflict and role ambiguity were related to being bullied and partially mediated
the relationship between a laissez-faire leadership style and bullying. Similar results
were obtained by Hauge, Skogstad, and Einarsen (2007), in which other work stressinducing factors such as job demands and decision authority, in addition to role
stressors, were related to being a victim of bullying, and by Notelaers, De Witte, and
Einarsen (2010), who found that role conflict and role ambiguity were the most
prominent correlates of being bullied. Furthermore, a recent study (Hauge, Skogstad,
& Einarsen, 2011) in which the analyses were conducted at the group level, largely
confirmed the same results, i.e. leadership practices and role conflict were

Work & Stress

197

departmental-level predictors of being a victim of bullying. Hodson, Roscigno, and


Lopez (2006), on the basis of an analysis of 148 organizational ethnographies,
concluded that a lack of coherence in production procedures creates chaotic work
environments in which bullying takes the place of more civil interactions. Bowling
and Beehr (2006), who meta-analysed the results in this area, concluded that role
stressors are the strongest potential antecedents of interpersonal conflict and
bullying. Evidence in line with this conclusion also emerges from the few studies
that have focused on bullying enactment. Hauge, Skogstad, and Einarsen (2009), for
example, found that role conflict significantly predicted the enactment of bullying.
Despite the convergence between the above-reported findings, most studies on
bullying have been based on cross-sectional data, so that a causal interpretation is
precluded. Only in a few cases has a longitudinal design been adopted. Baillien, De
Cuyper, and De Witte (2011) found a positive effect of job demands and lack of
autonomy on being the victim of bullying. Similarly, Baillien, Rodriguez-Munoz,
Van den Broeck, and De Witte (2011) found that a job demands factor including role
conflict, workload and job insecurity positively impacted on being bullied, while a
job resources factor consisting of skill utilization, autonomy and social support from
colleagues negatively affected the same outcome. However, Hauge, Skogstad, and
Einarsen (2011) found that role conflict, role ambiguity, and role overload did not
impact on being the victim of bullying longitudinally, while being bullied had a
positive longitudinal effect on all three dimensions. Overall, it may be concluded that
longitudinal evidence in favour of the work environment hypothesis of bullying is still
limited, and that there is a need for more research on the matter.
A further issue rarely considered when studying the role of the work environment
in bullying is the possible confounding effect of personal factors, which have been
repeatedly found to be involved in bullying (for a review see Zapf & Einarsen, 2011).
Personal factors may spuriously affect the relationship between work environmental
factors and bullying, especially in studies based exclusively on self-report  which is
the case of most studies in this area. Thus, to be more confident that the work
environment is indeed implicated in the escalation of bullying, a further important
step (see Cohen, Cohen, West, & Aiken, 2003; Zapf, Dorman, & Frese, 1996) is to
take potentially important personal factors into account.
It is of interest briefly to examine how personal factors may be involved in the
development of bullying. To be noted, however, is that although personal factors
may be related to both being a victim of bullying and its enactment (Zapf &
Einarsen, 2011), research in this area has mainly focused on the former outcome. In
this regard, Bowling, Beehr, Bennett, and Watson (2010) propose two mechanisms
whereby personal factors may contribute to bullying. The first suggests that
individuals with a chronically poor self-view may be easy victims for those who
enact bullying. The second posits that employees with certain personality traits such
as negative affectivity/neuroticism may behave in an irritating manner at work, which
may induce others to enact bullying towards them. Research has provided evidence
for both the mechanisms postulated by Bowling et al. (2010). In regard to the former
mechanism, studies have shown that a diagnosis of depression (Kivima ki et al., 2003)
and a measure of mental distress including symptoms of poor self-view (Finne,
Knardahl, & Lau, 2011) are longitudinal predictors of being bullied. In regard to the
latter mechanism, Bowling et al. (2010) found that negative affectivity has a
significant effect on workplace victimization. To be noted is that in the latter study

198

C. Balducci et al.

(Bowling et al., 2010), the authors controlled for the effect of work environmental
factors such as role stressors, which did not emerge as influencing the workplace
victimization. This suggests that personal psychological factors may even be the
principal causes of bullying.

The present study


As has recently been pointed out (Zapf & Einarsen, 2011), bullying research is still in
need of studies that conduct clear cause-effect analysis. In order to advance the
available knowledge on this topic, in the research reported here we used a
longitudinal design, and further investigated the potential antecedents of bullying
by focusing on the role of work environmental factors. We also controlled for
personal vulnerability, which  according to one of the mechanisms suggested by
Bowling et al. (2010)  may be involved in bullying, at least as far as being a victim of
the phenomenon is concerned.
Building upon previous evidence (Bowling & Beehr, 2006), we operationalized the
work environmental context of bullying in terms of role conflict and role ambiguity.
Role conflict refers to the simultaneous occurrence of two or more sets of pressures
on the focal person such that compliance with one makes compliance with the other
more difficult (Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964). Role ambiguity
refers to the inadequacy of role-related information, resulting in a lack of clarity
about duties, objectives, and responsibilities to fulfil ones role (Kahn et al., 1964). To
gain better understanding of why role stressors may be particularly important for the
development of bullying compared with other work environmental factors, we drew
on the work of Podsakoff, LePine, and LePine (2007), according to which job
stressors may be differentiated between challenge stressors and hindrance stressors.
Whilst all stressful job demands are subject to the same appraisal and coping
processes (Lazarus, 2006), the outcomes of these processes differ according to the
type of stressor being appraised. Challenge stressors, such as workload, time pressure
and responsibility, are generally appraised as opportunities, which implies that they
tend to evoke positive affective reactions and attitudes (Podsakoff et al., 2007). By
contrast, hindrance stressors, of which role ambiguity and role conflict are the
prototypical (and most studied) examples (see Podsakoff et al., 2007), are generally
appraised as potential threats to personal growth and goal attainment, which means
that they consistently tend to evoke negative emotions and attitudes. Thus, role
conflict and role ambiguity may be key factors in creating the work environmental
preconditions for the escalation of bullying, as suggested by the studies reviewed
above. To be noted further is that such deficiencies in the work role are prevalent
under certain conditions, such as when there are changes in technology and jobs, and
when there are shifts in economic stability and security threats (Beehr & Glazer,
2005); factors that have characterized the reality of work organizations in the past
decade or so (see Na swall, Hellgren, & Sverke, 2008).
To investigate the impact of role conflict and role ambiguity on bullying, we first
focused on being a victim of bullying and tested the following hypothesis:
Hypothesis 1. Exposure to role stressors (i.e. role conflict and role ambiguity) as
measured at time 1 (T1) affects being a victim of bullying at time 2 (T2) over and above
the effect of T1 personal vulnerability.

Work & Stress

199

Secondly, we also included the perspective of the perpetrators of bullying in the


analysis; and, in line with a frustration-aggression hypothesis underlying aggressive
behaviour at work (see Spector & Fox, 2005), we tested whether role conflict and role
ambiguity predicted the enactment of bullying. Specifically, the second hypothesis
was the following:
Hypothesis 2. Exposure to T1 role stressors (i.e. role conflict and role ambiguity)
impacts on the enactment of bullying at T2, over and above T1 personal vulnerability.

Since previous research has also found evidence that bullying impacts on work
environmental factors (Hauge et al., 2011), after testing each of the main study
hypotheses, we also explored reverse causation (e.g. that being bullied at T1 has a
lagged effect on role stressors at T2).

Method
Procedure and participants
Data were collected in two waves (at an interval of 12 months) from a sample of
healthcare workers employed in a National Health Service (NHS) agency in the north
of Italy. Data collection took place as part of a routine psychosocial risk assessment,
now mandatory in Italy for both private and public organizations. In agreement with
the organizations health and safety unit, the survey focused on those departments of
the organization that scored highest on a number of proxy indicators for work-related
stress (absences due to illness, turnover rates, disciplinary procedures, etc.). The data
for the study were collected by means of a self-reporting anonymous questionnaire,
which was administered to small groups of employees during working hours.
A total of 574 employees filled in the T1 questionnaire, with an average response
rate of 75.4% in the various departments. The T2 questionnaire was completed by
508 employees, with an average response rate of 65%. The T1 and T2 questionnaires
were matched by means of anonymous codes, which respondents built using factual
personal information. After deleting cases with incomplete data on the study
variables, follow-up data were available for 234 employees (40.1% of those filling in
the T1 questionnaire). Owing to the very sensitive nature of some of the data
collected, socio-demographic variables included in the questionnaires were formulated in such a way as to minimize the possibility of participant identification (e.g. for
the age variable, six broad age classes were formulated).
The following description of sample characteristics is based on T1 data; however, a
virtually identical description would be obtained with T2 data. Participants were
females in 86.3% of cases. The age classes most represented were 3039 years (37.3%)
and 4049 years (37.8%). The majority of participants were nurses (72.7%); also
represented were administrative staff (18.2%), medical doctors (4%), and other workers
(5.1%  e.g. cleaning personnel). Job tenure was more than five years for 83% of
participants. In most cases (97%), the employment contract was a permanent one.

Measures
Being a victim of bullying was investigated by using a shortened version (Notelaers &
Einarsen, 2008) of the Negative Acts Questionnaire (NAQ; Einarsen, Hoel, &

200

C. Balducci et al.

Notelaers, 2009), which has been validated in Italy (Balducci et al., 2010). This
version consists of nine items exploring how often the respondent has been subject to
a number of negative behaviours at work (e.g. Your work and effort have been
persistently criticized), with responses varying from 1 (Never) to 5 (Daily).
Previous research (Notelaers & Einarsen, 2008) has shown that the scale investigates
three closely correlated forms of bullying (i.e. work-related bullying, personal
bullying, and social isolation), and that it can be used as a single indicator of
exposure to the phenomenon. Cronbachs alpha for this scale and for the others
described below is reported in Table 1.
Bullying enactment was investigated by the nine items making up the bullying
measure (see above), which were rewritten in an active form (e.g. You have
persistently criticized the work and effort of someone at work). Responses varied
from 1 (Never) to 5 (Daily). These items derived from the NAQ have already been
used in previous research (e.g. Baillien, Rodriguez-Mun oz et al., 2011) to study
bullying behaviour from a perpetrators perspective. We acknowledge that there may
be some drawbacks in using such a measure  for example, its psychometric properties
have never been examined with an ad-hoc study. However, its important advantage
with respect to the other measures available, such as measures of interpersonal
deviance (Bennet & Robinson, 2000) or abuse/hostility counterproductive work
behaviour (Spector et al., 2006), is that it perfectly matches the bullying concept as
defined in the European tradition (see Zapf & Einarsen, 2005) and as it is captured  in
its passive operationalization  from a victims perspective. Cronbachs alpha for this
scale was slightly below the threshold (i.e. .68 at T1 and .66 at T2). However, it has
been argued that for less clearly-delimited psychological phenomena (of which
aggressive behaviour can be considered an example: see Spector et al., 2006),
measurement scales that attain an alpha of .60 to .70 can be regarded as acceptable
(Kline, 2000). It should also be noted that the alpha for this scale was similar to that
obtained in previous research (Baillien, De Cuyper, & De Witte, 2011).
Role conflict was measured by using five items (e.g. I receive incompatible
requests from two or more people) from the role conflict scale developed by Rizzo,
House, and Lirtzman (1970). Responses ranged from 1 (Entirely true) to 5 (Entirely
false), with items being reverse coded before the scale total was computed.
Role ambiguity was measured by using five items from a scale also developed by
Rizzo and colleagues (Rizzo et al., 1970), an example item being: I know what my
responsibilities are. The response format was the same as for the role conflict scale.
Organizational research (e.g. Kelloway & Barling, 1991) has supported the
separability of the role ambiguity and role conflict constructs as measured by the
Rizzo et al. (1970) scales. The scales have also been used in previous research in Italy
(e.g. Balducci, Fraccaroli, & Schaufeli, 2011).
Personal vulnerability was explored by using the 12-item version of the General
Health Questionnaire (GHQ-12; Goldberg, 1972). The GHQ-12 investigates the
respondents experience of a number of psychological symptoms (e.g. You have been
capable of making decisions), with responses ranging from 0 (No or More than
usual, according to specific items) to 3 (Much more than usual or Much less than
usual). The GHQ-12 has been well validated in Italy (e.g. Piccinelli, Bisoffi, Bon,
Cunico, & Tansella, 1993), with factor analytic studies (e.g. Claes & Fraccaroli, 2002)
suggesting the presence of three strongly-correlated factors: social dysfunction,
general dysphoria and loss of self-confidence. To identify participants with chronic

Table 1.

Properties and Pearsons product moment correlations of study variables (Cronbachs alpha on the diagonal, where appropriate).

Variable

SD

(.89)
.16*

(.68)

1.42 0.48
1.45 0.58
1.20 0.21

(.85)
.66**
.40**

1.31 0.30

.30**

.37**

.58**

0.43 0.50

.10

.17*

.05

.08

0.43 0.50

.16*

.25**

.04

.18**

.39**

0.05 0.22

.10

.15*

.18**

.10

.23**

.14*

0.06 0.23

.09

.14*

.01

.05

.21**

.17*

.62**

0.09 0.29

.10

.05

.04

.08

.29**

.16*

.39**

.37**

0.09 0.28

.02

-.01

.04

.10

.13

.20**

.36**

.41**

2.76
2.72
2.04
2.05
0.86
0.05

0.73 0.45

.04

0.83 0.38 .01


b

10

11

12

13

14

15

16

17

18

19

(.66)


0.87
.30**
.35**
.12**
.22** .17*
.17*
.11
0.86
.22**
.41**
.18**
.40** .07** .20** .07
0.59
.23**
.29**
.15*
.14
.15*
.18** .10
0.58
.28**
.48**
.12
.24** .17*
.21** .12
0.34 .08
.11
.04
.06
.11
.15*
.04
0.22 .05
.10
.10
.14* .04 .06 .01

0.18 0.39 .10

Work & Stress

1. T1 Being bullied
2. T2 Being bullied
3. T1 Bullying
enactment
4. T2 Bullying
enactment
5. T1 Psychological
caseness (GHQ-12)a
6. T2 Psychological
caseness (GHQ-12)
7. T1 Depressive
disorder diagnosis
8. T2 Depressive
disorder diagnosis
9. T1 Anxiety disorder
diagnosis
10. T2 Anxiety disorder
diagnosis
11. T1 Role conflict
12. T2 Role conflict
13. T1 Role ambiguity
14. T2 Role ambiguity
15. Genderb
16. Organizational role
(dummy 1)c
17. Organizational role
(dummy 2)d
18. Organizational role
(dummy 3)e
19. Tenuref


.55**

.07
.02
.04
.06
.04
.08

.10
.05
.10
.04
.09
.04

.05
.04
.08
.07
.12
.12

(.72)
.55**
(.75)
.44**
.29**
(.74)
.42**
.46**
.52**
(.77)
.02
.02
.01
.08
.06
.08
.16* .17*


.06

.01

.02

.07

.03

.03

.01

.06

.06

.04

.09

.10

.12

.10

.02

.07

.02

.03

.02

.03

.03

.01

.07

.04

.02

.02

.00

.02 .11

.07

.12

.09

.00

.05

.10

.11

.11

.05

.10

.02

.05

.04

.12 .18** .09

.09 .77**


.38**


.05 

Note: 0  GHQ-12 B 4, 1  GHQ-12 ] 4; 0  male; 1  females; 0  All others, 1  Administrative staff; 0  All others, 1  Nurses; 0  All others, 1  Other workers (e.g. cleaning staff);
f
0  5 years or less, 1  more than 5 years. *p B.05. **p B.01.

201

202

C. Balducci et al.

psychological problems, we used the so-called CGHQ scoring method (see Whaley,
Morrison, Payne, Fritschi, & Wall, 2005). Differently from the widely known
conventional scoring method, in which the same scoring algorithm (i.e. 0011) is
applied to all items, in the CGHQ method the response No more than usual to the
six items describing pathology is treated as an indicator of chronic problems and
scored accordingly (i.e. 1). We considered those participants with total GHQ-12
scores of four symptoms or more to be psychological cases (see Piccinelli et al., 1993).
In addition to the GHQ-12, personal vulnerability was also explored by means of the
following yes/no question: Have you received a medical diagnosis for any of the
following illnesses? The question was followed by a number of items, among which
depressive disorder and anxiety disorder featured.

Data analysis
Bullying is not an either/or phenomenon; rather, it is a gradually evolving process
(Einarsen et al., 2011) that may be best conceptualized as a continuum. It is highly
desirable to analyse such concepts as quantitative variables (Cohen et al., 2003).
Thus, to test the hypotheses proposed, we conducted two hierarchical regression
analyses. The dependent variable was being a victim of bullying at T2 in the first
analysis, and bullying enactment at T2 in the second. In both analyses, independent
variables were grouped into sets (see Cohen et al., 2003). Set 1 included only the T1
level of the dependent variable, and it was entered in the first step of the analysis. Set
2 included gender, T1 occupational position and T1 tenure, and it was added in the
second step. Set 3 included the three vulnerability variables considered (i.e.
psychological caseness, and self-reported medical diagnosis of depression and
anxiety disorder) as assessed at T1, and it was entered in the third step. Finally,
set 4 included the T1 role stress variables and was entered in step 4. The principles
underlying this hierarchical order were control of the baseline level of the dependent
variable and removal of confounding or spurious relationships.
Since the criterion variables were markedly positively skewed, in the interpretation of the results obtained we emphasized the role of the proportion of variance
accounted for (i.e. net contribution to prediction, see Cohen et al., 2003, p.152) by
each set of independent variables, which did not require the multivariate normality
assumption. This same interpretative strategy has already been adopted to
investigate bullying or similarly skewed variables (e.g. Baillien, Rodriguez-Mun oz
et al., 2011; Fox, Spector, & Miles, 2001; Marcus & Schuler, 2004).

Results
Before testing the main study hypotheses, we conducted an attrition analysis in order
to explore whether the variables on which the main analyses focused were related to
drop-out from the study, thus causing a selection bias (Cavallari, Eisen, Wegman, &
ONeill, 2011). We specifically conducted a logistic regression analysis on the 574
participants of T1, in which the outcome was participation vs. drop-out and the risk
factors were all the study variables: gender, occupational position and tenure, the
three vulnerability variables, role stressors and being bullied and bullying enactment.
We found that gender was strongly related to drop-out, with males being significantly

Work & Stress

203

more likely to leave the study than females (OR 2.5, pB.01). We found that also
being bullied was related to drop-out, with those participants more exposed to
bullying being significantly more likely to leave the study (OR 1.8, p B.05).
None of the other study variables was related to drop-out. The significant
relationship between being bullied and drop-out may have implications for
interpretation of the results of the main analyses. However, to be noted is that, for
selection bias to be a potential problem, drop-out should be related to both the
predictor(s) and the criterion (Cavallari et al., 2011, p. 521), which was not the case in
the present study.
Table 1 shows descriptive statistics, correlations and reliabilities for the study
variables. The mean values for the variables being bullied and bullying
enactment at both T1 and T2 fell between the response categories Never and
Now and then, indicating  as already anticipated  that the variables were quite
markedly positively skewed. These results are similar to those obtained by previous
studies in this area (e.g. Balducci, Fraccaroli, & Schaufeli, 2011; Hauge et al., 2011)
suggesting that most workers are not victims of (and do not enact) bullying. Of
interest is also the high stability of the being bullied (r.66) and bullying
enactment (r.58) measures, indicating a relatively small change in the bullying
situation between T1 and T2 for most of the participants. Furthermore, role conflict
and role ambiguity exhibited low to moderate correlations with being bullied, the
average correlation for both role stressors being r.32. The correlations between
role conflict and role ambiguity and bullying enactment were somewhat lower,
particularly in the case of role ambiguity (average r.29 for role conflict, and
average r .16 for role ambiguity).
Table 2 reports the results of the hierarchical regression analysis by which we
tested Hypothesis 1, i.e. that being bullied at T2 would be predicted by T1 role
stressors, even after controlling for T1 personal vulnerability. In light of the results of
attrition analysis as well as of previous research in this area (see, e.g. Baillien, De
Cuyper, & De Witte, 2011; Zapf, Escartn, Einarsen, Hoel, & Vartia, 2011) we also
controlled for gender, occupational position and tenure, as well as for being bullied at
T1. The results showed that the T1 personal vulnerability variables accounted for a
unique proportion of variance in being bullied at T2 over and above the T1 level of
the criterion variable (DR2 .025, pB.05), with a diagnosis of depressive disorder
contributing most to the increase in prediction (b .136, p B.05). The inclusion of T1
role stressors in step 4 of the model further increased its predictive power (DR2 .030,
pB.01), with role conflict (b .141, p B.05), rather than role ambiguity (b.074,
n.s.), playing a key role. These results provided some support for Hypothesis 1.
We also conducted parallel analyses (not reported in Table 1) to test the reverse
causation models. As regards role conflict, the results indicated that T1 role conflict
was the only variable explaining a significant proportion of variance in T2 role
conflict; however, to be noted is that being bullied at T1, which was included in the
last step of the equation after also controlling for T1 role ambiguity, explained a
further 1.3% variance in the criterion, a proportion that was almost significant
(p.056). As regards role ambiguity, after we had controlled for T1 level of the
construct, T1 role conflict made a significant contribution to the prediction
(DR2 .051, p B.01); the inclusion of the T1 being bullied variable in the last
step of the equation explained a further 1.5% variance in T2 role ambiguity, a
contribution that proved significant (pB.05).

204

C. Balducci et al.

Table 2. Hierarchical regression analysis predicting being bullied at T2 from role stressors at
T1 (N  234).
Predictors
T1 Being bullied
Gender
Organizational role (dummy 1)
(0  All others, 1  Administrative staff)
Organizational role (dummy 2)
(0  All others, 1  Nurses)
Organizational role 3 (dummy 3)
(0  All others, 1  Other workers,
e.g. cleaning staff)
Tenure (0  5 years or less, 1  more than 5 years)
T1 Psycological caseness
(0  GHQ-12 B 4, 1  GHQ-12 ] 4)
T1 Depressive disorder (0  No, 1  Yes)
T1 Anxiety disorder (0  No, 1  Yes)
T1 Role conflict
T1 Role ambiguity
DR2

Step 1 b Step 2 b Step 3 b Step 4 b


.66***

.43***

.66***
.65***
.58***
.00
.02
.02
.04
.04
.09
.01

.00

.05

.06

.06

.08

.06

.05
.08

.03
.06

.14*
.10

.13*
.10
.14*
.07
.03**

.01

.03*

Note: ***p B.001; **pB.01; *p B.05.

Table 3 reports the results of the hierarchical regression analysis by which we


tested Hypothesis 2, according to which T1 role stressors would impact on bullying
enactment at T2. T1 bullying enactment, socio-demographic variables and T1
vulnerability variables were also included to maintain the same structure of analysis
testing as used for Hypothesis 1. The results indicated that the last step of the
regression, in which we included T1 role conflict and T1 role ambiguity, accounted
for a unique and significant proportion of variance in the criterion (DR2 .030,
p.01). Again, it was T1 role conflict (b.20, pB.05), rather than T1 role
ambiguity (b .078, n.s.), that contributed most to this increase in prediction.
These results provided some support for Hypothesis 2. It should also be noted that,
contrary to the analysis focusing on being bullied (see above), in this case T1
vulnerability variables did not account for a significant proportion of variance in the
criterion variable. Parallel models were used to examine reverse causation. The
variance in T2 role conflict was not significantly predicted by the T1 level of bullying
enactment after controlling for T1 level of role conflict, socio-demographics, T1
personal vulnerability variables and T1 role ambiguity. As regards to T2 role
ambiguity, the only variable adding to its prediction over and above the T1 level of
role ambiguity was T1 role conflict (DR2 .054, pB.001).

Discussion
By using a longitudinal research design, we found that exposure to role stressors,
particularly role conflict, had a significant impact on being a victim of bullying one
year later, independent of the participants initial psychological vulnerability. In
other words, role stressors, and particularly situations such as being caught in the

Work & Stress

205

Table 3. Hierarchical regression analysis predicting bullying enactment at T2 from role


stressors at T1 (N  234).
Predictors
T1 Bullying enactment
Gender
Organizational role 1
(0  All others, 1  Administrative staff)
Organizational role 2
(0  All others, 1  Nurses)
Organizational role 3
(0  All others, 1  Other workers,
e.g. cleaning staff)
Tenure (0  5 years or less, 1  more than 5 years)
T1 Psychological caseness
(0  GHQ-12 B 4, 1  GHQ-12 ] 4)
T1 Depressive disorder (0  No, 1  Yes)
T1 Anxiety disorder (0  No, 1  Yes)
T1 Role conflict
T1 Role ambiguity
DR2

Step 1 b Step 2 b Step 3 b Step 4 b


.58***

.33***

.56***
.56***
.54***
.04
.03
.03
.28
.27
.32*
.21

.21

.23

.12

.12

.15

.05

.04
.03

.07
.02

.05
.11

.05
.10
.20**
.08
.03*

.02

.01

Note: ***p B.001; **pB.01; *p B.05.

crossfire of incompatible job demands, receiving an assignment without the adequate


resources and materials to execute it, working on unnecessary tasks, seem to play a
role as antecedents of being bullied. This is in line with the work environment
hypothesis on the escalation of bullying, according to which factors such as role
stressors, by triggering the stress process, may create the conditions for the
development of extreme and destructive forms of interpersonal conflicts such as
bullying. The effect of role stressors on being bullied was not very strong (accounting
for an additional 3.0% of its variance) and was mostly explained by role conflict.
However, this effect should not be underestimated when investigating complex and
multi-causal stress-related phenomena. According to Zapf et al. (1996), in stressorstrain research a set of variables (e.g. stressors) is unlikely to explain more than 45%
of variance in the strain phenomenon. Thus, since bullying may be conceptualized as
a multi-causal strain phenomenon (Einarsen et al., 2011), we expected role stressors
to explain at most a similar amount of its variance (i.e. 45%). To be noted is that,
according to Cohen et al. (2003, p. 151), even 12% of explained variance in the
criterion may represent a material effect, despite the customary disparagement of
effects of this magnitude.
Furthermore, although role conflict and role ambiguity are among the work
environmental factors that have been most investigated in relation to bullying, they
only represent two examples of hindrance stressors. There are other factors that tend
to act (at least predominantly) as hindrances, such as role overload, organizational
constraints and job insecurity. To be noted is that, for some of them, evidence of a
relationship with bullying has already been furnished (De Cuyper, Baillien, & de Witte,
2009). On considering a broader range of hindrance stressors, or perhaps by including
only those most prevalent in the organization investigated, it might be found that the
impact of the work environment on bullying increases. Furthermore, if the negative

206

C. Balducci et al.

emotional reactions generated by job stressors are indeed key factors in the
development of bullying, then job resources such as social support and autonomy,
by tending to promote positive job-related emotional experiences, should prevent its
escalation, further strengthening the influence of the work environment on bullying.
Finally, also to be considered is that perhaps even a small effect of work environmental
factors may be sufficient to trigger the initial stages of bullying; then, when
stigmatizing processes start and the victim becomes the problem (see Zapf &
Gross, 2001), bullying may fuel itself, so that being the target of victimizing behaviour
has a progressively stronger impact on undergoing further victimizing behaviour.
The effect of role stressors on bullying that emerged from our research has
particular significance as regards the validity of the work environment hypothesis.
This is because we controlled for participants personal vulnerability. In line with
previous research (e.g. Kivima ki et al., 2003), we found that participants initial
psychological functioning, particularly as revealed by the presence of a diagnosis of
depression, had an effect on being a victim of bullying, which supports the view that
psychological vulnerability may signal to potential perpetrators that the person is an
easy target. The effect of role stressors over and above personal vulnerability suggests,
as an avenue for future research, the adoption of a personenvironment perspective in
which certain psychological traits/problems may be particularly important for
attracting bullying behaviour under poor working conditions  i.e. when potential
perpetrators may experience those critical internal states (e.g. high arousal) that are
the immediate preconditions for bullying enactment. This perspective, which has been
rarely adopted in this area (for an example, see Balducci, Fraccaroli, & Schaufeli,
2011), may refine and further increase our understanding of bullying.
We would emphasize that the results of this study are compatible with a view of a
reciprocal influence between role stressors and bullying. Indeed, we found that being
bullied affected the level of role ambiguity one year later. Furthermore, the effect of
being bullied on subsequent role conflict was almost significant. This means that
being bullied led to a certain intensification of the experience of role stressors. This is
not an original finding (see Hauge et al., 2011). Furthermore, being a victim of
bullying by definition generates marked role problems, since some bullying actions
(e.g. removal of areas of responsibility, assignment of meaningless tasks) may literally
destroy the victims work role, a situation that is captured, in its less extreme aspects,
by the role stressor constructs. What is more important to understand, especially for
preventive purposes, is whether initial working conditions may trigger the bullying
process. A similar reciprocal relationship most probably exists between the
experience of bullying and personality. Exposure to long-lasting and extreme
bullying may alter the victims personal characteristics. This was one of the initial
findings of bullying research (see Leymann & Gustafsson, 1996). What is important
to understand, again for preventive purposes, is whether initial personality factors
may contribute to becoming a victim of bullying.
In line with our second hypothesis, we also found that T1 role stressors impacted
on T2 bullying enactment. It was particularly role conflict that contributed to this
effect. This finding confirms, from a different standpoint, what has emerged from the
perspective of victims, and provides further evidence for the involvement of the work
environment in the escalation of bullying. The importance of role stress specifically
due to role conflict as an antecedent of bullying is strengthened. This result seems to
be particularly compatible with the idea that role stressors may act as ambient

Work & Stress

207

stressors, leading to the bullying of some individuals because they trigger aggressive
behaviour on the part of potential perpetrators, an explanation that is gaining
consensus among bullying researchers (see the theoretical model recently proposed
by Einarsen et al., 2011, p. 29). Of particular interest is that, on testing for the effect
of role stress on bullying enactment, we found that those vulnerability variables that
were important for being bullied only made a very marginal contribution to bullying
enactment. This supports the notion that victims and perpetrators may have at least
partially different personal characteristics (see Zapf & Einarsen, 2011): for example,
victims may have depressive and withdrawal tendencies, while perpetrators may have
high but unstable self-esteem or narcissistic tendencies. Other personal characteristics
that may be important for the enactment of bullying may be trait anger, effortful
control, and hostile attribution bias (see Spector, 2011). Unfortunately we did not
assess any of these, which may be considered a shortcoming of the present study.
Main methodological issues and limitations
There are a number of methodological issues that may need to be considered in
relation to the contribution made by this study. Firstly, it was conducted in the health
sector, which seems to be a risk sector as far as bullying and intimidation are
concerned (Zapf et al., 2011). Secondly, we concentrated on those organizational
departments that were most at risk of work-related stress according to certain proxy
measures of strain, such as absenteeism due to illness, turnover, or disciplinary
procedures. Hence, although the study sample had a certain approximation to the
parent population for some characteristics  for example, females constitute the large
majority of the workers in the Italian health service, with the proportion reaching
70% in northern regions (Italian Ministry of Health, 2011)  we do not know
whether the findings can be generalized to other types of organization.
However, we note that the focus on a specific subgroup of employees may also be a
strength of the study as far as the understanding of bullying escalation is concerned.
Taris and Kompier (2003), for example, suggest that, in order to gain better
understanding of stress-related phenomena, occupational health research would
benefit from focusing on theoretically relevant subgroups of participants, such as those
exposed to the more adverse working conditions or those experiencing higher levels of
strain. This is because strain phenomena (of which bullying may be an example)
usually affect a minority of employees, so that including all the employees of an
organization in the analyses may dilute the magnitude of already small relationships
between the phenomena of interest. In this regard, it is interesting to note that in the
research reported by Hauge et al. (2011), which was based on a representative sample
of the Norwegian population, and in which the same measure of exposure to bullying
was used, the mean level of reported negative acts was significantly lower at the t-test
than that reported by the present study at both T1 and T2. This may be a possible
reason for the lack of effect by role stressors on bullying that emerged in their study.
A further issue to consider is that bullying in the health sector may be enacted not
only by work colleagues or superiors, but also by patients and/or clients (European
Foundation, 2007). We do not know whether the victims of bullying in our study
were reporting negative behaviours from other members versus non-members of the
organization. Although this could be an interesting research question for the purpose
of developing a more refined view of bullying, it has been a secondary issue for the

208

C. Balducci et al.

present study, since its main interest has been investigation of the relationship
between the quality of the work environment and bullying. Furthermore, when
investigating bullying enactment, we found that some employees did indeed engage in
bullying, meaning that there is at least some correspondence between what was
reported by victims and perpetrators in terms of prevalence of the phenomenon.
Another issue that warrants discussion is the one-year time lag used in the
present study to test the relationship between role stressors and bullying. The time
lag considered is a critical element in the investigation of the relationship between
stressors and strain phenomena (De Lange, Taris, Kompier, Houtman, & Bongers,
2004; Taris, 2000; Zapf et al., 1996). If the adopted time lag does not match the
underlying causal mechanism, the estimation of the true relationship between the
variables of interest may be biased (Taris, 2000). Previous longitudinal research on
the antecedents of bullying has not been consistent with regard to the time lag
chosen, and the results supported different causal mechanisms. With time lags of six
months (Baillien, de Cuyper, & De Witte, 2011) and one year (Baillien, RodriguezMun oz et al., 2011), evidence emerged for an impact of work environmental factors
on bullying, while evidence for reverse causation did not. By contrast, with a time lag
of two years (Hauge et al., 2011) only evidence for reverse causation (i.e. bullying
impacting on the quality of the work environment) was obtained. In the present
study, in which we used a time lag of one year, we found an effect of work
environmental factors on bullying, but also some evidence for reverse causation. The
overall picture suggested by these results is that the effect of the work environment on
bullying seems to be the only causal mechanism involved when adopting relatively
short time lags (i.e. six months). On lengthening the observation period to one year,
the effect of the work environment on bullying is apparently the predominant one,
but some evidence of a reciprocal relationship seems to emerge. With longer time lags
(i.e. two years), reverse causation becomes the only evident causal mechanism,
suggesting an effect of bullying on the quality of the work environment. It may well
be that a poor work environment tends to escalate bullying in a relatively short
period. However, as bullying develops it may start to influence the work environment
negatively and cause its quality to deteriorate further. Moreover, since bullying is an
escalating process that fuels itself (Zapf & Gross, 2001) and has pervasive effects on
the work organization, its influence may become progressively stronger, so that in the
longer run the initial triggering effect of work environmental factors is no longer
observable. Future research should pay closer attention to the issue of the time lag
used in investigating the antecedents of bullying. There is perhaps a need for
theoretical as well empirical work on this aspect.
An additional issue is that the variables used in the analyses were all selfreported, which means that their relationships may have been affected by common
method bias (Chan, 2009). Furthermore, particularly in the case of bullying
enactment, under-reporting due to social desirability may have occurred. However,
the main hypotheses of the study were tested by means of longitudinal analysis,
which is the measure recommended to limit the effects of common method bias
(Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). As far as the possible underreporting of bullying enactment is concerned, this may have led to an underestimation of the relationship between bullying enactment and the other variables
(see Chan, 2009), but will not have affected the main results of the study.

Work & Stress

209

Finally, there is the issue of the distribution of the focused dependent variables (i.e.
being bullied and bullying enactment), which was non-normal. We used hierarchical
linear regression and particularly focused on the percentage of explained variance in
the criteria, which does not require the normality assumption (Cohen et al., 2003).
However, significance tests might have been biased. An alternative option with nonnormal data would be structural equation modelling, which allows for robust
estimation methods (see Joreskog & So rbom, 1996). However, the number of estimated
parameters may easily become high, so that large samples are generally required.
Despite the above issues and limitations, we believe that this study has some
strengths that others in this area generally lack. Specifically, it has been based on a
longitudinal design (one-year time interval between the two waves of data collection);
it has considered both work environmental and personal (vulnerability) factors; and
it has included the perspectives of both victims and perpetrators of bullying.
In terms of implications, the results of this study strengthen the idea that problems
of poor work organization should be addressed in order to prevent the escalation of
extreme and destructive forms of interpersonal conflict such as bullying (Vartia &
Leka, 2011). However, since personal factors seem also to be involved in bullying,
secondary and perhaps tertiary preventive interventions are necessary as well.

Concluding remarks
In line with the work environment hypothesis of bullying (Einarsen et al., 1994;
Leymann, 1996) we have found that role stressors and particularly role conflict may
lead to the escalation of bullying. However, we have also found some evidence for
reverse causation, suggesting a reciprocal relationship between working conditions
and bullying. We hope that this study will prompt further, methodologically robust
studies on bullying development, which are much needed in this area of research.
References
Aquino, K., & Thau, S. (2009). Workplace victimization: Aggression from the targets
perspective. Annual Review of Psychology, 60, 717741.
Baillien, E., De Cuyper, N., & De Witte, H. (2011). Job autonomy and workload as
antecedents of workplace bullying: A two-wave test of Karaseks Job Demand Control
Model for targets and perpetrators. Journal of Occupational and Organizational Psychology,
84, 191208.
Baillien, E., Neyens, I., De Witte, H., & de Cuyper, N. (2009). A qualitative study on the
development of workplace bullying: Towards a three way model. Journal of Community and
Applied Social Psychology, 19, 116.
Baillien, E., Rodriguez-Mun oz, A., Van den Broeck, A., & De Witte, H. (2011). Do demands
and resources affect targets and perpetrators reports of workplace bullying? A two-wave
cross-lagged study. Work & Stress, 25, 128146.
Balducci, C., Alfano, V., & Fraccaroli, F. (2009). Relationships between mobbing at work and
MMPI-2 personality profile, post traumatic stress symptoms and suicidal ideation and
behaviour. Violence and Victims, 24(1), 5267.
Balducci, C., Fraccaroli, F., & Schaufeli, W. (2011). Workplace bullying and its relation with
work characteristics, personality, and post-traumatic stress symptoms: An integrated
model. Anxiety, Stress & Coping, 24(5), 499512.

210

C. Balducci et al.

Balducci, C., Schaufeli, W.B., & Fraccaroli, F. (2011). The Job Demands-Resources model and
counterproductive work behaviour: The role of job-related affect. European Journal of Work
and Organizational Psychology, 20(4), 467496.
Balducci C., Spagnoli P., Alfano V., Barattucci M., Notelaers G., & Fraccaroli F. (2010).
Valutare il rischio mobbing nelle organizzazioni: Contributo alla validazione italiana dello
Short Negative Acts Questionnaire (S-NAQ) [Assessing the mobbing risk in organizations:
Contribution to the Italian validation of the Short Negative Acts Questionnaire (S-NAQ)].
Psicologia Sociale, 1, 147167.
Beehr, T.A., & Glazer, S. (2005). Organizational role stress. In J. Barling, K. Kelloway, &
M. Frone (Eds.), Handbook of work stress (pp. 733). Thousand Oaks, CA: Sage.
Bennet, R.J., & Robinson, S.L. (2000). Development of a measure of workplace deviance.
Journal of Applied Psychology, 85(3), 349360.
Berkowitz, L. (1989). Frustration-aggression hypothesis: Examination and reformulation.
Psychological Bulletin, 106(1), 5973.
Bowling, N.A., & Beehr, T.A. (2006). Workplace harassment from the victims perspective:
A theoretical model and meta-analysis. Journal of Applied Psychology, 9(5), 9981112.
Bowling, N.A., Beehr, T.A., Bennett, M.M., & Watson, C.P. (2010). Target personality and
workplace victimization: A prospective analysis. Work & Stress, 24, 140158.
Cavallari, J.M., Eisen, E.A., Wegman, D.H., & ONeill, M.S. (2011). Epidemiology. In B.S.
Levy, D.H. Wegman, S.L. Baron, & R.K. Soaks (Eds.), Occupational and environmental
health (pp. 507526). New York, NY: Oxford University Press.
Chan, D. (2009). So why ask me? Are self-report data really that bad? In C.E. Lance & R.J.
Vandenberg (Eds.), Statistical and methodological myths and urban legends (pp. 309336).
New York, NY: Routledge.
Claes, R., & Fraccaroli, F. (2002). Il General health Questionnaire (GHQ-12): Invarianza
fattoriale in diverse versioni linguistiche [The General Health Questionnaire (GHQ-12):
Factorial invariance in different linguistic versions]. Bollettino di Psicologia Applicata, 237,
2535.
Cohen, J., Cohen, P., West, S.G., & Aiken, L.S. (2003). Applied multiple regression/correlation
analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates.
De Cuyper, N., Baillien, E., & De Witte, H. (2009). Job insecurity and workplace bullying
among targets and perpetrators: Moderation by employability. Work & Stress, 23, 20622.
De Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D., & Bongers, P.M. (2004). The
relationships between work characteristics and mental health: Examining normal, reversed
and reciprocal relationships in a 4-wave study. Work & Stress, 18, 142166.
Einarsen, S., Hoel, H., & Notelaers, G. (2009). Measuring exposure to bullying and
harassment at work: Validity, factor structure and psychometric properties of the Negative
Acts Questionnaire-Revised. Work & Stress, 23, 2444.
Einarsen, S., Hoel, H., Zapf, D., & Cooper, C.L. (2011). The concept of bullying and
harassment at work: The European tradition. In S. Einarsen, H. Hoel, D. Zapf, & C.L.
Cooper (Eds.), Bullying and harassment in the workplace. Developments in theory, research,
and practice (pp. 339). Boca Raton, FL: CRC Press.
Einarsen, S., Raknes, B.I., & Matthiesen, S.B. (1994). Bullying and harassment at work and
their relationship with to work environment quality: An exploratory study. European Work
and Organisational Psychologist, 4, 381401.
European Foundation [for the Improvement of Living and Working Conditions] (2007).
Fourth European working conditions survey. Luxembourg: Office for Official Publications of
the European Communities.
Finne, L.B., Knardahl, S., & Lau, B. (2011). Workplace bullying and mental distress  A
prospective study of Norwegian employees. Scandinavian Journal of Work Environment &
Health, 37(4), 27686.
Fox, S., Spector, P.E., & Miles, D. (2001). Counterproductive work behavior (CWB) in
response to job stressors and organizational justice: Some mediator and moderator tests for
autonomy and emotions. Journal of Vocational Behavior, 59, 291309.

Work & Stress

211

Goldberg, D.P. (1972). The detection of psychiatric illness by questionnaire. Maudsley


monograph N. 21. London: Oxford University Press.
Hauge, L.J., Einarsen, S., Knardhal, S., Lau, B., Notelaers, G., & Skogstad, A. (2011). Leadership
and role stressors as departmental level predictors of workplace bullying. International journal of
Stress Management, 18, 305323.
Hauge, L.J., Skogstad, A., & Einarsen, S. (2007). Relationships between stressful work
environments and bullying: Results of a large representative study. Work & Stress, 21, 220242.
Hauge, L.J., Skogstad, A., & Einarsen, S. (2009). Individual and situational predictors of
workplace bullying: Why do perpetrators engage in the bullying of others? Work & Stress,
23, 349358.
Hauge, L.J., Skogstad, A., & Einarsen, S. (2011). Role stressors and exposure to workplace
bullying: Causes and consequences of what and why? European Journal of Work and
Organizational Psychology, 20, 610630.
Hodson, R., Roscigno, V.J., & Lopez, S.H. (2006). Chaos and the abuse of power:
Workplace bullying in organizational and interactional context. Work and Occupations,
33(4), 382416.
Hogh, A., Mikkelsen, E.G., & Hansen, A.M. (2011). Individual consequences of workplace
bullying/mobbing. In S. Einarsen, H. Hoel, D. Zapf & C.L. Cooper (Eds.), Bullying and
harassment in the workplace. Developments in theory, research, and practice (pp. 107128).
Boca Raton, FL: CRC Press.
Italian Ministry of Health (2011). Women in the Italian national healthcare service. Retrieved
from: http://www.salute.gov.it/dettaglio/phPrimoPianoNew.jsp?id 304
Jo reskog, K., & So rbom, D. (1996). LISREL 8 users guide. Chicago, IL: Scientific Software.
Kahn, R.L., Wolfe, D.M., Quinn, R.P., Snoek, D.J., & Rosenthal, R.A. (1964). Organizational
stress: Studies in role conflict and ambiguity. New York, NY: Wiley.
Kelloway, E.K., & Barling, J. (1991). Item content versus item wording: Disentangling role
conflict and role ambiguity. Journal of Applied Psychology, 75(6), 738742.
Kivima ki, M., Virtanen, M., Vartia, M., Elovainio, M., Vahtera, J., & Keltikangas-Ja rvinen, L.
(2003). Workplace bullying and the risk of cardiovascular disease and depression.
Occupational and Environmental Medicine, 60, 779783.
Kline, P. (2000). The handbook of psychological testing (2nd ed). London: Routledge.
Lazarus, R.S. (2006). Stress and emotions: A new synthesis. New York, NY: Springer.
Leymann, H. (1996). The content and development of mobbing at work. European Journal of
Work and Organizational Psychology, 5, 165184.
Leymann, H., & Gustafsson, A. (1996). Mobbing at work and the development of posttraumatic stress disorders. European Journal of Work and Organizational Psychology, 5(2),
251275.
Marcus, B., & Schuler, H. (2004). Antecedents of counterproductive behavior at work:
A general perspective. Journal of Applied Psychology, 89(4), 647660.
Mikkelsen, E.G., & Einarsen, S. (2002). Basic assumptions and symptoms of post-traumatic
stress among victims of bullying at work. European Journal of Work and Organizational
Psychology, 11(1), 87111.
Na swall, K., Hellgren, J., & Sverke, M. (Eds.) (2008). The individual in the changing working
life. New York, NY: Cambridge University Press.
Neuman, J.H., & Baron, R.A. (2011). Social antecedents of bullying: A social interactionist
perspective. In S. Einarsen, H. Hoel, D. Zapf, & C.L. Cooper (Eds.), Bullying and
harassment in the workplace. Developments in theory, research, and practice (pp. 201225).
Boca Raton, FL: CRC Press.
Notelaers, G., De Witte, H., & Einarsen, S. (2010). A job characteristics approach to
explain workplace bullying. European Journal of Work and Organizational Psychology, 19,
478495.
Notelaers, G., & Einarsen, S. (2008, June). The construction and validity of the Short  Negative
Acts Questionnaire. Paper presented at the 6th International Conference on Bullying and
Harassment in the Workplace. Montreal, Canada.

212

C. Balducci et al.

Piccinelli, M., Bisoffi, G., Bon, M.G., Cunico, L., & Tansella, M. (1993). Validity and testretest reliability of the Italian version of the 12-item General Health Questionnaire in
general practice: A comparison between three scoring methods. Comprehensive Psychiatry,
34(3), 198205.
Podsakoff, N.P., LePine, J.A., & LePine, J.A. (2007). Differentisal challenge stressor-hindrance
stressor relationships with job attitudes, turnover intentions, and withdrawal behavior: A
meta-analysis. Journal of Applied Psychology, 92(2), 438454.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., & Podsakoff, N.P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended remedies.
Journal of Applied Psychology, 88, 879903.
Rizzo, J.R., House, R.J., & Lirtzman, S.I. (1970). Role conflict and ambiguity in complex
organizations. Administrative Science Quarterly, 15(2), 150163.
Skogstad, A., Einarsen, S., Torsheim, T., Aasland, M.S., & Hetland, H. (2007). The
destructiveness of laissez-faire leadership behavior. Journal of Occupational Health
Psychology, 12(1), 8092.
Spector, P. (2011). The relationship of personality to counterproductive work behavior (CWB):
An integration of perspectives. Human Resource Management Review, 21, 342352.
Spector, P.E., & Fox, S. (2005). The stressor-emotion model of counterproductive work
behaviour. In S. Fox & P.E. Spector (Eds.), Counterproductive work behaviour (pp. 151174).
Washington, DC: American Psychological Association.
Spector, P.E., Fox, S., Penney, L.M., Bruursema, K., Goh, A., & Kessler, S. (2006). The
dimensionality of counterproductivity: Are all counterproductive behaviors created equal?
Journal of Vocational Behavior, 68, 446460.
Taris, T.W. (2000). A primer in longitudinal data analysis. London: Sage.
Taris, T.W., & Kompier, M. (2003). Challenges in longitudinal designs in occupational health
Psychology. Scandinavian Journal of Work, Environment and Health, 29(1), 14.
Vartia, M., & Leka, S. (2011). Interventions for the prevention and management of bullying at
work. In S. Einarsen, H. Hoel, D. Zapf, & C.L. Cooper (Eds.), Bullying and harassment in
the workplace. Developments in theory, research, and practice (pp. 359379). Boca Raton, FL:
CRC Press.
Whaley, C.J., Morrison, D.L., Payne, R.L., Fritschi, L., & Wall, T.D. (2005). Chronicity of
psychological strain in occupational settings and the accuracy of the General Health
Questionnaire. Journal of Occupational Health Psychology, 10(4), 310319.
Zapf, D., Dorman, C., & Frese, M. (1996). Longitudinal studies in organizational stress
research: A review of the literature with reference to methodological issues. Journal of
Occupational Health Psychology, 1(2), 145169.
Zapf, D., & Einarsen, S. (2005). Mobbing at work: escalated conflicts in organizations. In
S. Fox & P.E. Spector (Eds.), Counterproductive work behaviour (pp. 237270). Washington,
DC: American Psychological Association.
Zapf, D., & Einarsen, S. (2011). Individual antecedents of bullying: Victims and perpetrators.
In S. Einarsen, H. Hoel, D. Zapf, & C.L. Cooper (Eds.), Bullying and harassment in the
workplace. Developments in theory, research, and practice (pp. 177200). Boca Raton, FL:
CRC Press.
Zapf, D., Escartn, J., Einarsen, S., Hoel, H., & Vartia, M. (2011). Empirical findings on
prevalence and risk groups of bullying in the workplace. In S. Einarsen, H. Hoel, D. Zapf, &
C.L. Cooper (Eds.), Bullying and harassment in the workplace. Developments in theory,
research, and practice (pp. 75105). Boca Raton, FL: CRC Press.
Zapf, D., & Gross, C. (2001). Conflict escalation and coping with workplace bullying:
A replication and extension. European Journal of Work and Organizational Psychology,
10(4), 497522.

Copyright of Work & Stress is the property of Taylor & Francis Ltd and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.

You might also like