Workplace Harassment From The Victim's Perspective: A Theoretical Model and Meta-Analysis

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/6835933
Workplace Harassment from the Victim's Perspective: A Theoretical Model

and Meta-Analysis
Article in Journal of Applied Psychology · October 2006

DOI: 10.1037/0021-9010.91.5.998 · Source: PubMed
CITATIONS READS
613 12,043
2 authors, including:
Nathan Bowling
Wright State University
69 PUBLICATIONS 3,099 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Your attention please! Toward a better understanding of research participant carelessness View project
Research and Theory on Workplace Aggression View project
All content following this page was uploaded by Nathan Bowling on 07 February 2015.
The user has requested enhancement of the downloaded file.

Journal of Applied Psychology Copyright 2006 by the American Psychological Association
2006, Vol. 91, No. 5, 998 –1012 0021-9010/06/$12.00 DOI: 10.1037/0021-9010.91.5.998
Workplace Harassment From the Victim’s Perspective: A Theoretical

Model and Meta-Analysis
Nathan A. Bowling and Terry A. Beehr
Central Michigan University
Although workplace harassment affects the lives of many employees, until recently it has been relatively
ignored in the organizational psychology literature. First, the authors introduced an attribution- and
reciprocity-based model that explains the link between harassment and its potential causes and conse-
quences. The authors then conducted a meta-analysis to examine the potential antecedents and conse-
quences of workplace harassment. As shown by the meta-analysis, both environmental and individual
difference factors potentially contributed to harassment and harassment was negatively related to the
well-being of both individual employees and their employing organizations. Furthermore, harassment
contributed to the variance in many outcomes, even after controlling for 2 of the most commonly studied
occupational stressors, role ambiguity and role conflict.
Keywords: bullying, harassment, workplace aggression, interpersonal conflict, occupational health psy-
chology
The social environment at work can have both positive and individual versus aimed at the organization (Robinson & Bennett,
negative effects on people. Among the most serious negative 1995).
effects are those caused by workplace harassment, which is de-
fined as interpersonal behavior aimed at intentionally harming A Model of a Victim’s View of Workplace Harassment
another employee in the workplace, and it represents “a burgeon-
ing research area in organizational psychology” (Aquino & Figure 1 proposes a victim’s view of workplace harassment (see
Lamertz, 2004, p. 1023). Research on workplace harassment ap- Box 5 in the figure). The variables in the figure are numbered for
pears under many different labels (e.g., bullying, interpersonal easy reference. The model has underlying themes of reciprocity
conflict, social undermining, and abuse) and in different disci- and attribution processes (Kelley, 1973). Research in many do-
plines (Aquino & Lamertz, 2004; Einarsen, 2000; Keashly & mains illustrates powerful effects of the norm of reciprocity (Cial-
Jagatic, 2000), but each label refers to the same overall construct dini, 2001; Gouldner, 1960). Victims might reciprocate with reac-
(Aquino & Lamertz, 2004; Lapierre, Spector, & Leck, 2005). The tions directed toward the harassment perpetrator or toward the
present meta-analysis brings coherence to this research by sum- organization if they attribute the cause of harassment to them.
marizing and organizing it around theory. Attribution for the cause of harassment can easily be made to the
Extreme harassment can include homicide and physical assault, specific perpetrator, but there are also multiple reasons to blame
but the more common, minor instances include obscene gestures, (an attribution process) the organization because it can be seen as
dirty looks, threats, yelling, giving the silent treatment, and belit- responsible for the actions and even the very presence of the
tling. The type of harassment reviewed here is not motivated by the perpetrator. Self-attributions are not expected to lead to reciprocal
sex or race of the victim (Einarsen, Raknes & Matthiesen, 1994) exchange with the organization or the perpetrator, however. In this
and is thus not covered by Title VII of U.S. civil rights law. Theft case, harassment is experienced primarily in terms of stress, and
from the organization, showing up late for meetings, and whistle typical stress outcomes are likely to occur.
blowing (Neuman & Baron, 1997) can also be considered forms of Victims’ personal characteristics might affect harassment, but
aggression, but they do not fit the present definition of harassment victims are also likely to see the organization’s climate and human
because they are directly aimed at or harm the organization rather resources systems (selection, training, and reward systems) as
than being aimed at a specific person. This is similar to the partially responsible through their effects on the nature of the
distinction between counterproductive work behavior aimed at the workplace (presence of other stressors and presence of people who
harass). Causal attributions can lead to the victim’s experience of
poor well-being, to attitudes and actions toward the perpetrator,
and to perceptions of organizational injustice. If victims blame
Nathan A. Bowling and Terry A. Beehr, Department of Psychology, themselves, poor well-being, such as reduced self-esteem and
Central Michigan University.
increased depression, are likely to result. Negative attitudes and
We thank Bryan Gibson and Stephen H. Wagner for their helpful
comments regarding earlier versions of this article. We also thank the many
actions toward the perpetrator, in contrast, can be a form of
authors who graciously shared their unpublished research with us. reciprocity due to attributing causation to the perpetrator; indeed,
Correspondence concerning this article should be addressed to Nathan A. Aquino and Lamertz (2004) reviewed research showing that per-
Bowling, who is now at the Department of Psychology, Wright State Univer- petrators tend to report frequently being victims as well (e.g., r ⫽
sity, Fawcett 313F, Dayton, OH 45435. E-mail: psybowling@yahoo.com .52 between being a victim and a perpetrator; Glomb & Liao,
998
META-ANALYSIS OF WORKPLACE HARASSMENT 999
(1) Organizational Culture/Climate
(2) Organizational Human Resource

Systems
(3) Perpetrator (4) Victim

Characteristics Characteristics
(5) Workplace Harassment (6) Role

Stressors
(7) Number (8) Process of

of Victims Attribution to:
and
(9)
Perpetrators
(8a) Self Victim’s
Well-Being
(11a) (10) Attitudes

(8b) Perpetrator Organizational and Behaviors
Injustice toward
(Interactional) Perpetrator
(11b) (12)
(8c) Organization
Organizational Individual
Injustice Performance
(Distributive and Outcomes
Procedural)
Figure 1. Reciprocity and attribution processes in workplace harassment. Boxes are numbered for ease of
reference with the text.
2003). Finally, perceptions of organizational injustice lead to neg- Antecedents of Harassment: Roles of Organization,
ative reactions that the victim justifies as retaliation against an Individual Differences, and Other Stressors
organization that allows harassment.
Theoretically, there are at least three categories of causes for
People tend to make causal attributions to others for unpleasant
workplace harassment: characteristics of the work environment,
events such as harassment, just as they do for the unpleasant
the perpetrator, and the victim. Victims can attribute or blame two
experience of their own poor performance (Shaw, Wild, &
of these antecedents on the organization. First, the organization
Colquitt, 2003; Zuckerman, 1979), because a self-attribution for
can be seen as directly responsible for the presence of a perpetrator
being harassed would reflect badly on one’s self. In Figure 1, the
because its human resources systems (Box 2) select, train, and
victim’s perception of the numbers of perpetrators and victims reward or punish perpetrators. Second, an organizational climate
(Box 7) in the organization is one guiding factor in making or culture (Box 1) permitting or encouraging harassment can be
attributions: If there are many victims and perpetrators, then the judged responsible both directly and indirectly (Aquino &
organization can be seen as responsible; if there are many victims Lamertz, 2004). In addition to climate’s direct effects on harass-
but only one perpetrator, then the attribution can easily be made to ment (Box 5), it can foster other forms of stress (Box 6), and
the perpetrator because harassment is unique to that person; if the harassment can flourish in stressful climates. Victims can easily
person is the only victim and there are multiple perpetrators, then blame the organization for its climate and for the existence of
the victim could blame him- or herself; and if the victim sees only certain role stressors.
one victim and one perpetrator in the organization, then it is a Role stressors (e.g., role conflict and ambiguity; Box 6) are
unique dyadic situation (see Aquino & Lamertz, 2004, for descrip- related to workplace harassment (e.g., Spector & O’Connell,
tions) and the attribution is indeterminate— however, an ego- 1994), perhaps because harassers experience stressful emotions
protective tendency might prevail and lead to one of the attribu- and lash out at coworkers. Harassment and role stressors have
tions to others. some commonality. Both are inherently social; role stressors are
1000 BOWLING AND BEEHR
characteristics of messages from other people at work, and harass- In addition to dispositional variables, some studies have exam-
ment consists of behaviors from the same set of people. ined the role of victims’ demographic characteristics. One possi-
There are a number of reasons why stressors might be anteced- bility is that victim demographics could influence the effect–
ents of harassment. First, being exposed to work stressors might danger ratio (Bjorkqvist, Osterman, & Lagerspetz, 1994). That is,
produce negative emotional and behavioral responses that encour- perpetrators’ judgments of the potential costs and benefits of
age victimization. Second, victims who report the presence of high harassing could be influenced by victim demographics because
levels of stressors might be indicating the presence of ambient certain demographic groups (e.g., women or those with low tenure)
stressors experienced by everyone in the victim’s work environ- would tend to be “safer” targets. The empirical research in this area
ment. Thus, when victims experience stressors, perpetrators might is mixed, however. Some research, for example, found that women
be experiencing many of the same stressors. High levels of stres- are more likely to become the targets of workplace harassment
sors could cause some perpetrators to engage in harassment. than men are (e.g., Bjorkqvist, Osterman, & Hjelt-Back, 1994;
In addition to climate and job stress, personal characteristics can Cortina, Magley, Williams, & Langhout, 2001), but others found
help determine potential harassment perpetrators. Researchers sug- no gender differences (e.g., Einarsen & Skogstad, 1996; Leymann,
gest that perpetrator (Box 3) impulsivity, emotional reactivity, and 1996). In addition, the relationship between victim age and bully-
rebelliousness (Andersson & Pearson, 1999); cynicism and low ing seems to differ greatly across studies (Hoel et al., 1999; Rayner
tolerance for ambiguity (Ashforth, 1997); gender (e.g., Zapf, Ein- & Hoel, 1997). Overall, although environmental and personal
arsen, Hoel, & Vartia, 2003); and hierarchical position (e.g., Ein- characteristics have been examined as potential antecedents of
arsen & Skogstad, 1996; Rayner, 1998) are related to becoming a workplace harassment, there is a need to get a more clear reading
perpetrator. Greenberg and Barling (1999) found positive correla- of the major trends.
tions between aggression toward subordinates, coworkers, and Rather than separate characteristics of the perpetrator and the
supervisors, suggesting the cause might reside in the perpetrator. If victim, their joint characteristics might be responsible for harass-
perpetrator characteristics are partly responsible for the harassment. Regarding gender, for example, one study suggested that
ment, then victims can logically blame either the perpetrator or the male bullies target both male and female victims but female bullies
organization (for allowing the harasser’s behavior). generally target only women (Rayner & Hoel, 1997). Observers
Victims who experience conflict at work also experience con- often note there are likely to be power differences between per-
flict with their spouses, family, and friends (Skjorshammer & petrator and victim (Bjorkqvist et al., 1994; Einarsen, 2000; Ein-
Hofoss, 1999), and this broad experience of conflict suggests some arsen et al., 1994; Einarsen & Skogstad, 1996; Hoel et al., 1999).
characteristics of victims (Box 4) could also be antecedents to Types of power differences might include greater physical size, a
harassment. This would not lead to reciprocity. Indeed, victims more aggressive personality, a more powerful formal position in
might tend to believe themselves relatively unresponsible for ha- the organization, or numbers (more people on his or her side).
rassment; however, it would be incomplete to ignore their role Effects at the intersection of characteristics of the perpetrator and
(Aquino & Lamertz, 2004; Lamertz & Aquino, 2004). Leymann victim are an interesting theme, but little systematic empirical
(1996) argued that a victim’s personality does not predict harass- research exists on them.
ment, but other reviewers concluded the relationship is unclear Overall, therefore, individual differences are antecedent factors
(e.g., Hoel, Rayner, & Cooper, 1999). Researchers might be re- in the incidence of workplace harassment. At the organization
luctant to examine this topic for fear that results could reinforce a level, however, the organization’s climate and human resources
tendency to blame the victim, but several studies report victims’ systems are proposed to have an effect on the incidence of harass-
traits are linked with harassment. ment, both directly and indirectly.
Positive affectivity (PA) and negative affectivity (NA) of the
victim have been examined, with victims’ PA negatively and NA
positively associated with workplace harassment (e.g., Aquino, Consequences of Harassment: Roles of Attribution and
Grover, Bradfield, & Allen, 1999; Duffy, Ganster, & Pagon, 2002; Reciprocity
Spector & O’Connell, 1994; Tepper, Duffy, Hoobler, & Ensley,
2006), but results were sometimes nonsignificant (e.g., Duffy et Theoretical and empirical work has identified negative conse-
al., 2002; Tepper, Duffy, & Shaw, 2001). Thus, although there is quences for both victims’ and organizations’ well-being (for re-
a strong research theme regarding these two particular traits, views, see Einarsen & Mikkelsen, 2003; Hoel, Einarsen, & Coo-
results are not completely consistent. Other than NA and PA, per, 2003; Keashly & Jagatic, 2000). Some of the ultimate
Aquino (2000) found that victims’ submissiveness was, but vic- consequences in Figure 1 are victims’ reactions that can be seen as
tim’s domination was not, associated with being bullied. Clearly, reciprocal. Probably the most direct and obvious instances of
the research on victim personality characteristics is far from com- reciprocity involve negative reactions toward the perpetrator (Box
plete. Effects of other characteristics are also logical. Among the 10), which occur when the attribution is that the perpetrator him-
Big Five (McCrae & Costa, 1987) personality characteristics, for or herself is mainly responsible for the harassment (Box 8b). There
example, agreeableness and conscientiousness are two possibili- can also be reciprocity in negative reactions toward the organiza-
ties. People who are conscientious and agreeable would seem less tion (Boxes 11b and 12), however; that is, there are negative
likely to be a target of harassment because they do less to irritate responses toward the organization in response to the organization
coworkers who are potential perpetrators than would people who allowing the harassment to occur. Other responses are directed at
are disagreeable and undependable. Although people who are more the self and resemble effects of work-related stress (Box 9) rather
irritating can partly lead to victimization, research lags behind this than reciprocity. Both reciprocity and stress outcomes are pro-
explanatory theme. posed to be mediated by attribution processes (Box 8).
In occupational research, aspects of one’s work environment posed to result from attributions of causation to the organization
that could be harmful to the employee and require adaptive re- (Box 8c), and the process of reciprocity then leads to lowered
sponses are often known as stressors (e.g., role conflict, role individual performance outcomes (Box 12), which include both in-
ambiguity), and the negative health consequences of those stres- and extrarole performance as well as withdrawal from and lower
sors are called strains (e.g., depression, physical symptoms; Jex, commitment to the organization.
Beehr, & Roberts, 1992). Workplace harassment is one example of
a stressor (Spector & Jex, 1998), and several strains have been
Workplace Harassment as a Stressor Compared With
investigated in relation to harassment. Figure 1 proposes this
occurs primarily when victims attribute blame for the harassment
Other Stressors
to themselves (Box 8a). Indeed, the majority of victims and wit- The potential consequences of workplace harassment are similar
nesses of bullying report experiencing at least some negative to those of other stressors, such as role ambiguity and role conflict
health and well-being (Box 9) consequences (Rayner, 1998), a sign (Box 6). As noted earlier, both harassment and role stressors are
harassment is a stressor. social phenomena, and it is logical to ask whether their effects are
It is easy to understand that the experience of harassment could actually independent or are redundant with each other. Whereas
harm one’s self-image because being picked on implies low status decades of research examined the effects of role ambiguity and
and powerlessness. Seen this way, harassment is a form of stressor role conflict (Beehr & Glazer, 2005; Cooper & Dewe, 2004),
in the category of stress as offense to self (Semmer, McGrath, & research has only relatively recently begun to examine the stress
Beehr, 2005). A few studies examined the relationship between effects of workplace harassment. Thus, it is logical to ask whether
workplace harassment and self-esteem (e.g., Ashforth, 1997; workplace harassment has effects that go beyond these more
Frone, 2000), self-efficacy (e.g., Mikkelsen & Einarsen, 2002), or traditional, better-known stressors. We examined this question by
self-confidence (e.g., Vartia, 2001), but not all research actually applying hierarchical regression analysis.
found a relationship between workplace harassment and self-
esteem (e.g., Duffy et al., 2002; Zapf, Knorz, & Kulla, 1996).
Theories of reciprocity in the form of employee retaliation The Meta-Analysis
(Skarlicki & Folger, 1997) or revenge (Bies & Tripp, 2005)
Workplace harassment, interpersonal behavior aimed at in-
suggest that when employees are harassed at work they are likely
tentionally harming another employee in the workplace, has
to experience feelings of injustice, resentment, and frustration that
garnered increasing interest among organizational researchers.
could result in reciprocation in the form of behaviors aimed at
It has important implications for the person being harassed and
“getting even” with the organization. A norm of reciprocity is
for the employing organization. The present study is a meta-
pervasive in many social activities and can have a powerful influ-
analysis of its correlates. The model in Figure 1 proposes
ence on human behavior (Cialdini, 2001; Gouldner, 1960). The
antecedents and consequences of workplace harassment, cate-
work organization is a place where the reciprocity norm is espe-
gorized as the victim is likely to perceive the situation. Both
cially strong, given that a primary reason for people’s presence
systemic organizational properties and individual difference
there is to exchange skills and effort for pay and respect. There-
characteristics are likely to influence the experience of harass-
fore, reciprocal exchanges are expected, common, and considered
ment. Regarding the consequences of harassment, the model
appropriate. In Figure 1, the employees’ final reactions to the
relies heavily on theories of attribution and reciprocity combin-
effects of harassment can be considered here as reciprocity reac-
ing to explain, along with stress principles, many of the effects
tions; they complete a set of (negative) exchanges with the orga-
of harassment. Although there is good rationale for the model,
nization. In spite of the likelihood of these organizationally rele-
data are not yet available to conduct a meta-analysis on all parts
vant reactions to the experience of workplace harassment,
of the model or to test it as a complete structural model. Much
however, fewer studies have examined the potential effects of
research has, however, been conducted on direct links between
harassment on organizational welfare than have examined the
harassment and most of the variables in the study. We examine
effects on individual welfare. Consistent with this, several years
this research as partial evidence regarding the model. Because
ago in a small meta-analysis (k ⫽ 2), Spector and Jex (1998) found
previous meta-analyses were available on some of the relation-
that interpersonal conflict correlated –.10 with job performance.
ships between role stressors (Box 6 in the model) and victims’
Figure 1 indicates that harassment leads to reciprocity through
well-being (Box 4 in the model), we were able also to test this
organizational justice perceptions (Box 11). Poor treatment by the
particular mediated link by combining results gathered from
organization or its members can lead to feelings of injustice
those previous meta-analyses of occupational stress with results
(Colquitt, Conlon, Wesson, Porter, & Ng, 2001), and feelings of
of the present meta-analysis of harassment.
injustice have been linked to workplace harassment in a few
studies (e.g., Cortina et al., 2001; Frone, 1998b; Tepper, 2000).
The form of injustice that one experiences likely depends on Method
whether victims attribute the cause of the harassment to them-
Potential antecedents and consequences of workplace harassment were
selves, to the perpetrator, or to the organization. Injustice is not
subjected to meta-analysis. The PsycINFO computer database identified
experienced when victims attribute the harassment to themselves. workplace harassment research published from 1987 through 2005. The
In contrast, distributive and procedural injustice is experienced search terms were abuse, abusive supervision, aggression, bullying, ha-
when the harassment is attributed to the organization, and inter- rassment, incivility, interpersonal conflict, mistreatment, mobbing, petty
actional injustice is experienced when the harassment is attributed tyranny, and social undermining. The reference sections of the articles
to the perpetrator. Perceptions of organizational injustice are pro- identified during this step were reviewed to find additional articles. Au-
thors who had recently published research on workplace harassment were Table 1
also contacted to request copies of unpublished manuscripts. Average Internal Consistency Reliabilities for All Study
These methods identified 168 journal articles, books, book chapters, Variables
unpublished manuscripts, theses, and dissertations. Some of these were
either nonempirical or did not report any of the relationships examined by Variable Reliability (␣)
the meta-analyses. For example, Andersson and Pearson (1999) presented
a theoretical basis for studying workplace incivility and Ashforth (1994) Harassment .81
Role conflict .80
provided a conceptual discussion of the antecedents and consequences of
Role ambiguity .82
petty tyranny. Such articles were excluded from the meta-analyses. Other Role overload .80
studies (e.g., Budd, Arvey, & Lawless, 1996; Rospenda, Richman, Wislar, Work constraints .87
& Flaherty, 2000) failed to report statistics necessary to conduct the Autonomy .82
meta-analyses, such as a correlation coefficient, N, F ratio, exact p, or t Positive affectivity .79
statistic, and thus they were excluded. A number of studies (e.g., Baldwin, Negative affectivity .86
Daugherty, & Eckenfels, 1991; Richman, Flaherty, Rospenda, & Chris- Gender 1.00
Age 1.00
tensen, 1992; Sheehan, Sheehan, White, Leibowitz, & Baldwin, 1990;
Tenure 1.00
Silver & Glicken, 1990; Wolf, Randall, Von Almen, & Tynes, 1991) Generic strain .87
examined medical student harassment rather than workplace harassment. Anxiety .86
These studies were also excluded. Depression .86
A total of 90 samples were included in the meta-analyses. Twenty-three Burnout .84
of these samples came from unpublished sources (8 conference papers, 11 Frustration .73
dissertations, 3 theses, and 1 manuscript submitted for publication). Each Positive emotions at work .90
Negative emotions at work .88
was a field study with data provided by employed adults. Several different
Self-esteem .84
measures of harassment scales were represented in the current meta- Life satisfaction .88
analysis. The most common harassment measures were the Interpersonal Physical symptoms .80
Conflict at Work Scale (ICAWS; Spector & Jex, 1998; 28 samples), the Job satisfaction .83
Negative Acts Questionnaire (Einarsen & Raknes, 1997; eight samples), Organizational commitment .84
and the Abusive Supervision Scale (Tepper, 2000; six samples). All studies Organizational justice .90
examined the antecedents and/or consequences of harassment from self- Job performance .83
Organizational citizenship behavior .89
report data provided by would-be victims rather than would-be perpetrators Counterproductive work behavior .87
or third parties. Studies that measured any combination of victim-reported Turnover intention .83
verbal, psychological, or physical harassment at work were selected for Absenteeism 1.00
inclusion in the meta-analysis, but data involving vicarious harassment
were excluded. A few studies used multiple measures of harassment (e.g., Note. Variables listed as having an alpha of 1.00 were assessed with
separate measures for verbal and physical harassment) or included separate single-item measures.
measures of harassment that corresponded to different groups of perpetra-
tors (e.g., supervisors, coworkers, or customers). In such cases, the corre-
lations for the various measures were averaged to create a single studywise Hierarchical Regression Analysis Regarding
correlation for harassment. Consequences
A growing number of studies have applied regression analysis to meta-
Meta-Analytic Strategy analytically derived correlation matrices (Viswesvaran & Ones, 1995). In
the current study, we used hierarchical regression analyses to examine
Only some parts of the model, derived theoretically in the introduction whether workplace harassment was related to employee attitudes, strains,
(see Figure 1), have actually been tested by previous empirical research— and behaviors after role ambiguity and role conflict were controlled. This
was done because workplace harassment, as a stressor, tended to occur
those parts involving role stressors, perpetrator and victim characteristics,
when other stressors were also present and data obtained from previous
organizational justice, victim’s well-being, and individual performance
meta-analyses on role conflict and role ambiguity allowed us to test the
outcomes. The rest of the model, including the attribution processes and the
incremental effects of harassment. These analyses used the parameter
organizational variables that would be necessary for a clear test of reci-
estimates obtained in the current meta-analysis as well as those obtained in
procity effects, needs future research.
a prior meta-analysis on role conflict and ambiguity (Collins, 1999; Craw-
We used Hunter and Schmidt’s (2004) meta-analysis method. Zero-order
ford, 2000; Jackson & Schuler, 1985). We were thus limited to examining
correlations, sample sizes, and, when available, reliability data for predic- only those outcomes that had been previously addressed by the meta-
tors and criteria were used to conduct the meta-analysis. Average reliability analysis on the stressors of role conflict and role ambiguity.
data for each variable can be found in Table 1. If an article reported a First, we examined a regression model that included the traditional
measure of effect size other than correlations (e.g., F ratio or t value), stressors of role ambiguity and role conflict as predictor variables. We then
equations provided by Hunter and Schmidt were used to transform the added the newer stressor, workplace harassment, to the model, observing
statistic into a correlation coefficient. the magnitude of its beta and the change in the percentage of variance
Correlations across samples were averaged, weighting each observed explained. The sample size for each analysis was the harmonic mean of the
correlation by the study’s sample size, and each correlation was corrected sample sizes reported in the contributing meta-analyses (Viswesvaran &
for unreliability in the predictor and criterion. Artifact distributions were Ones, 1995).
used to estimate the value of missing reliability data; otherwise, the studies We also examined the extent to which workplace harassment mediated
were individually corrected for unreliability. Credibility intervals and per- the relationships between role ambiguity and role conflict on one hand and
centage of variance in effect sizes explained by artifacts were computed the outcome variables on the other. In these analyses, we computed the
using the methods suggested by Hunter and Schmidt (2004). effect proportion mediated (PM; Shrout & Bolger, 2002), which is a ratio
reflecting the extent to which the total effects of X on Y are explained by it was negatively associated with positive emotions at work (␳ ⫽
the indirect (i.e., mediated) effects of X on Y. –.25), self-esteem (␳ ⫽ –.21), life satisfaction (␳ ⫽ –.21), job
satisfaction (␳ ⫽ –.39), and organizational commitment (␳ ⫽
Moderator Analysis –.36). Furthermore, as shown in Table 4, harassment was nega-
tively associated with perceptions of organizational justice (␳ ⫽
We conducted moderator analysis to examine whether the relationships
between harassment and its potential antecedents and consequences dif- –.35).
fered depending on whether the ICAWS (Spector & Jex, 1998) was used In contrast, as shown in Table 4, workplace harassment was
to assess harassment. More specifically, we were interested in examining associated with some individual performance outcomes but not
whether the results obtained in the above analyses were influenced by the with others. In particular, harassment was positively associated
overrepresentation of studies using the ICAWS. This was important, as with counterproductive work behaviors (␳ ⫽ .37) and turnover
mentioned above, because 28 studies included in the meta-analysis used the intentions (␳ ⫽ .35), but it was weakly related to job performance
ICAWS to measure harassment. We made no a priori predictions concern- (␳ ⫽ –.08), organizational citizenship behaviors (␳ ⫽ –.03), and
ing the results of these moderator analyses. absenteeism (␳ ⫽ .06).
Results
Hierarchical Regression Analysis
Results of the meta-analyses on the antecedents and conse-
quences of workplace harassment are in Tables 2, 3, and 4. Table 5 presents the results of the regression analyses. The
analyses presented in the table used corrected correlations. We also
Potential Antecedents of Harassment conducted these analyses using uncorrected correlations, which
yielded similar results as those presented in the table.
As shown in Table 2, workplace harassment tends to occur in
As shown in Table 5, both role ambiguity (␤ ⫽ .14) and role
work environments where other stressors are present. Specifically,
conflict (␤ ⫽ .38) were uniquely related to harassment. Together,
some other stressors, such as role conflict (␳ ⫽ .44), role ambiguity
these two stressors predicted 21% of the variance in workplace
(␳ ⫽ .30), role overload (␳ ⫽ .28), and work constraints (␳ ⫽ .53),
harassment. The hierarchical regression analyses, which are also
were all associated with workplace harassment. Autonomy (␳ ⫽
depicted in the table, indicated that workplace harassment was
–.25) was another feature of the work environment associated with
harassment. related to a number of outcomes even after role ambiguity and role
Victim NA appears to play some role in workplace harassment conflict were controlled. Specifically, harassment predicted prac-
(␳ ⫽ .25). In contrast, victim PA was only weakly related to tically significant amounts of incremental variance in burnout (R2
harassment (␳ ⫽ –.09), as were victim demographic variables, change ⫽ .06), physical symptoms (R2 change ⫽ .05), job satis-
such as gender (male ⫽ 1, female ⫽ 2; ␳ ⫽ –.05), age (␳ ⫽ –.04), faction (R2 change ⫽ .03), organizational commitment (R2
and tenure (␳ ⫽ .02). change ⫽ .04), and turnover intention (R2 change ⫽ .04).
Because the prior meta-analyses (Crawford, 2000; Jackson &
Schuler, 1985) used to conduct these regression analyses examined
Potential Consequences of Harassment
role ambiguity’s but not role conflict’s relationships with depres-
As shown in Table 3, workplace harassment was associated with sion and self-esteem, the regression analyses for these two out-
victims’ well-being. Specifically, harassment was positively asso- come variables only controlled for role ambiguity. As shown in
ciated with generic strains (␳ ⫽ .35), anxiety (␳ ⫽ .31), depression Table 5, workplace harassment predicted practically significant
(␳ ⫽ .34), burnout (␳ ⫽ .39), frustration (␳ ⫽ .40), negative incremental variance in depression (R2 change ⫽ .04) after role
emotions at work (␳ ⫽ .46), and physical symptoms (␳ ⫽ .31), and ambiguity was controlled.
Table 2
Relationships Between Workplace Harassment and Antecedents
% variance 80% credibility

Variable k N Mean r Mean ␳ SD␳ accounted for interval
Role conflict 16 5,429 .35 .44 .08 37.39 .34, .55

Role ambiguity 22 6,759 .24 .30 .12 23.32 .14, .46
Role overload 25 7,343 .22 .28 .06 52.50 .19, .37
Work constraints 13 2,733 .44 .53 .14 18.64 .35, .72
Autonomy 13 2,823 ⫺.20 ⫺.25 .06 58.80 ⫺.34,⫺.16
Victim positive affectivity 8 2,293 ⫺.08 ⫺.09 .09 37.96 ⫺.21, .02
Victim negative affectivity 24 7,441 .21 .25 .14 17.41 .07, .44
Victim gender 11 2,921 ⫺.05 ⫺.05 .00 100.00 ⫺.05,⫺.05
Victim age 16 4,822 ⫺.04 ⫺.04 .08 38.65 ⫺.14, .05
Victim tenure 13 4,504 .01 .02 .02 88.75 .00, .04
Note. For gender, male ⫽ 1 and female ⫽ 2. k refers to the number of samples, N refers to the total sample size, mean r refers to the average weighted
correlation coefficient, mean ␳ refers to the average weighted correlation coefficient corrected for unreliability in both the predictor and criterion, and %
variance accounted for refers to the variance accounted for by artifacts.
Table 3
Relationships Between Workplace Harassment and Consequences: Victims’ Well-Being

Generic strain 27 17,663 .29 .35 .08 23.00 .24, .46

Anxiety 16 4,918 .25 .31 .11 26.29 .17, .45
Depression 16 5,625 .28 .34 .12 18.81 .18, .50
Burnout 9 5,633 .33 .39 .16 7.10 .18, .61
Frustration 16 3,613 .30 .40 .08 48.91 .29, .51
Positive emotions at work 4 1,783 ⫺.21 ⫺.25 .03 65.73 ⫺.30,⫺.20
Negative emotions at work 7 1,549 .38 .46 .13 21.48 .28, .64
Self-esteem 9 3,066 ⫺.17 ⫺.21 .14 14.94 ⫺.40,⫺.02
Life satisfaction 3 1,851 ⫺.18 ⫺.21 .00 100.00 ⫺.21,⫺.21
Physical symptoms 33 13,878 .25 .31 .11 23.00 .17, .46
Job satisfaction 42 19,871 ⫺.32 ⫺.39 .10 22.70 ⫺.52,⫺.25
Organizational commitment 16 9,224 ⫺.30 ⫺.36 .15 9.42 ⫺.55,⫺.16
Note. Generic strain measures assessed multiple strains (e.g., depression and physical symptoms) in a single scale. k refers to the number of samples, N
refers to the total sample size, mean r refers to the average weighted correlation coefficient, mean ␳ refers to the average weighted correlation coefficient
corrected for unreliability in both the predictor and criterion, and % variance accounted for refers to the variance accounted for by artifacts.
Table 5 also presents the PM (Shrout & Bolger, 2002), which those variables where the k for both the ICAWS studies and the
examines the extent to which workplace harassment mediated the non-ICAWS studies was four or more.
effects of role ambiguity and role conflict on the outcome vari- In these analyses, we examined practical rather than statistical
ables. The results of these analyses indicated that in general, the significance. We did this because with large Ns, differences in effect
indirect effects of role ambiguity and role conflict on the outcome sizes that are trivial by practical standards may yield statistical sig-
variables accounted for only a small proportion of the total effects nificance. For the present moderator analyses, we judged the differ-
of role ambiguity and role conflict on the outcome variables. The ence between effect sizes for a given variable to be practically sig-
largest PM, for example, was .42 (the relationship between role nificant if differences in percentages of variance explained (␳2)
ambiguity and physical symptoms). Thus, the indirect effects of between ICAWS and non-ICAWS measures were 5% or greater.
role ambiguity on physical symptoms that were mediated through For several variables, the effect sizes did not yield any practical
workplace harassment explained 42% of the total effects of role differences across type of harassment measure used (see Table 6).
ambiguity on physical symptoms. Many of the other PMs included Specifically, harassment yielded similar relationships with role
in Table 5 were much closer to .00, suggesting that overall the overload, gender, age, tenure, generic strains, physical symptoms,
direct effects were stronger than the indirect effects.
job satisfaction, and turnover intention regardless of whether the
ICAWS or some other measure was used to assess harassment.
The ICAWS as a Moderator Studies using the ICAWS did, however, yield larger corrected
The moderator analyses compared the effect sizes of studies correlations than did studies not using the ICAWS for role conflict
using the ICAWS (Spector & Jex, 1998) with those of studies not (␳ ⫽ .52 vs. ␳ ⫽ .36), role ambiguity (␳ ⫽ .36 vs. ␳ ⫽ .22), NA
using the ICAWS to measure harassment. These moderator anal- (␳ ⫽ .35 vs. ␳ ⫽ .20), anxiety (␳ ⫽ .46 vs. ␳ ⫽ .25), and depression
yses were not practical for some of the variables in Tables 2, 3, and (␳ ⫽ .42 vs. ␳ ⫽ .28). Autonomy, in contrast, was the only variable
4 because grouping studies on the basis of whether they used the for which non-ICAWS studies yielded an average corrected cor-
ICAWS would yield a very small number of samples (k) for the relation that was practically greater than that of the ICAWS studies
moderator analyses. We thus conducted moderator analyses on (␳ ⫽ –.34 vs. ␳ ⫽ –.20).
Table 4
Relationships Between Workplace Harassment and Consequences: Organizational Justice and Individual Performance Outcomes

Organizational justice 8 2,061 ⫺.29 ⫺.35 .13 23.22 ⫺.51,⫺.18

Job performance 5 976 ⫺.06 ⫺.08 .00 100.00 ⫺.08,⫺.08
Organizational citizenship behavior 5 1,493 ⫺.02 ⫺.03 .18 12.43 ⫺.26, .19
Counterproductive work behavior 9 2,584 .30 .37 .06 53.18 .29, .46
Turnover intention 24 13,961 .29 .35 .11 18.93 .21, .49
Absenteeism 5 1,267 .06 .06 .06 52.91 ⫺.01, .15
Note. k refers to the number of samples, N refers to the total sample size, mean r refers to the average weighted correlation coefficient, mean ␳ refers to
the average weighted correlation coefficient corrected for unreliability in both the predictor and criterion, and % variance accounted for refers to the variance
accounted for by artifacts.
Table 5
Incremental Validity of Workplace Harassment Controlling for Other Stressors (Role Ambiguity and Role Conflict)
Effect proportion
Dependent variable Predictor N ␤ R2 change mediated (PM)
Harassment Role ambiguity 6,996 .14** .21**

Role conflict .38**
Anxiety Role ambiguity 5,921 .34** .29** .02
Role conflict .24** .00** .14
Workplace harassment .10**
Depression Role ambiguity 1,289 .45** .26** .11
Workplace harassment .21** .04**
Burnout Role ambiguity 5,070 .17** .16** .19
Physical symptoms Role ambiguity 947 .04 .07** .42
Self-esteem Role ambiguity 3,672 ⫺.30** .12** .11
Workplace harassment ⫺.12** .01**
Job satisfaction Role ambiguity 8,173 ⫺.29** .31** .09
Role conflict ⫺.28** .03** .20
Workplace harassment ⫺.18**
Organizational commitment Role ambiguity 4,718 ⫺.28** .21** .09
Role conflict ⫺.15** .04** .34
Workplace harassment ⫺.21**
Turnover intention Role ambiguity 4,898 .15** .14** .16
Absence Role ambiguity 1,256 .16** .02** .05
Role conflict ⫺.11** .01* .22
Workplace harassment .06*
Note. N is the harmonic mean of the individual sample sizes for the parameters that contribute to each analysis. All betas are from the final regression
model after workplace harassment was added. The first change in R2 reflects the percentage of variance in the dependent variable explained on the first
step of the regression equation. The second change in R2 reflects the incremental variance in the dependent variable explained by adding workplace
harassment to the regression equation.
* p ⬍ .05. ** p ⬍ .01.
Discussion predispose employees to engage in behaviors that provoke mis-

treatment from others. Although each of these possibilities is
The meta-analysis yielded some important findings concerning
interesting, the current meta-analysis cannot address them directly
the potential causes and consequences of workplace harassment. It
and they thus need to be examined in future empirical studies.
appears that characteristics of the work environment (e.g., other
It is also possible that NA is a consequence of workplace
stressors) might strongly contribute to workplace harassment. This
harassment. Being continually abused could certainly have an
finding can be interpreted in a number of ways. First, employees
effect on one’s personality. This might especially be true among
who work in stressful work environments may have a tendency to
individuals exposed to extremely severe or chronic forms of ha-
become victims of harassment because the stressors in their jobs
rassment. It is likewise possible that self-esteem could be either an
produce behavioral and affective reactions that encourage others to
victimize them. Second, individuals who report the presence of antecedent or consequence of harassment. Unfortunately, little
other stressors could potentially be reporting a negative ambient longitudinal research has been conducted to examine the causal
work environment that would-be perpetrators are also experienc- relationships between workplace harassment on one hand and NA
ing. Such negative work environments might encourage would-be and self-esteem on the other.
perpetrators to engage in harassing behaviors. Overall, this might In addition to NA, we would expect other personality charac-
suggest a climate in which harassment is common. teristics of victims to play a role. Conscientiousness and agree-
Compared with the environmental antecedents, victims’ individ- ableness, for example, might both be negatively associated with
ual differences (dispositional and demographic characteristics) harassment. Supervisors should have better relationships with con-
seem to have little influence on whether an employee is harassed. scientious subordinates, and thus conscientiousness would be neg-
Victims’ NA was an exception, and it could influence harassment atively associated with harassment from supervisors. Furthermore,
in a number of ways. For example, NA might predispose individ- agreeableness is likely to be negatively related to harassment
uals to perceive the presence of harassment regardless of whether because agreeable employees are more pleasant to be around and
harassment has actually occurred. Other possibilities are that NA are less likely to engage in behaviors that provoke harassment.
influences the tendency for employees to self-select into work Subordinate agreeableness and conscientiousness might be seen by
environments where harassment is already common or it might supervisors as indicators of competence; rather than being ha-
Table 6
Moderator Analyses by Harassment Scale
% variance 80%
accounted credibility
Variable Scale k N Mean r Mean ␳ SD␳ for interval
Role conflict ICAWS 12 2,729 .41 .52 .00 100.00 (.52, .52)
Other 4 2,700 .30 .36 .08 21.19 (.25, .47)
Role ambiguity ICAWS 17 3,864 .29 .36 .05 62.57 (.29, .44)
Other 5 2,895 .18 .22 .14 9.83 (.03, .40)
Role overload ICAWS 19 5,835 .22 .29 .06 55.99 (.21, .37)
Other 6 1,508 .19 .24 .08 44.36 (.13, .35)
Autonomy ICAWS 9 1,878 ⫺.17 ⫺.20 .09 45.41 (⫺.31,⫺.08)
Other 4 945 ⫺.25 ⫺.34 .00 100.00 (⫺.34,⫺.34)
Negative affectivity ICAWS 10 2,605 .28 .35 .07 46.85 (.25, .44)
Other 14 4,836 .18 .20 .15 13.46 (.01, .39)
Gender ICAWS 5 1,251 ⫺.06 ⫺.07 .00 100.00 (⫺.07,⫺.07)
Other 6 1,670 ⫺.04 ⫺.04 .00 100.00 (⫺.04,⫺.04)
Age ICAWS 6 1,033 .01 .01 .05 74.17 (⫺.05, .08)
Other 10 3,789 ⫺.05 ⫺.05 .08 31.73 (⫺.16, .05)
Tenure ICAWS 4 817 .05 .06 .00 100.00 (.06, .06)
Other 9 3,687 .01 .01 .01 92.89 (.00, .02)
Generic strain ICAWS 5 2,905 .32 .38 .02 76.43 (.35, .41)
Other 22 14,758 .29 .34 .09 20.38 (.22, .46)
Anxiety ICAWS 7 1,244 .37 .46 .00 100.00 (.46, .46)
Other 9 3,674 .22 .25 .09 27.29 (.13, .37)
Depression ICAWS 8 2,244 .35 .42 .09 30.08 (.29, .54)
Other 8 3,381 .24 .28 .11 16.23 (.13, .43)
Physical symptoms ICAWS 17 3,737 .21 .29 .07 56.83 (.19, .39)
Other 16 10,141 .26 .31 .11 14.29 (.16, .46)
Job satisfaction ICAWS 20 6,162 ⫺.29 ⫺.35 .00 97.82 (⫺.36,⫺.34)
Other 22 13,709 ⫺.33 ⫺.39 .12 13.84 (⫺.54,⫺.23)
Turnover intention ICAWS 10 2,358 .31 .37 .11 29.49 (.23, .51)
Other 14 11,603 .28 .35 .10 16.57 (.21, .49)
Note. For gender, male ⫽ 1 and female ⫽ 2. k refers to the number of samples, N refers to the total sample size, mean r refers to the average weighted
correlation coefficient, mean ␳ refers to the average weighted correlation coefficient corrected for unreliability in both the predictor and criterion, and %
variance accounted for refers to the variance accounted for by artifacts. ICAWS ⫽ Interpersonal Conflict at Work Scale (Spector & Jex, 1998).
rassed by supervisors, agreeable and conscientious subordinates The Model

are likely to be part of the supervisor’s favored in-group, in terms
of leader–member exchange theory (Gerstner & Day, 1997). Un- Although only parts of the model have been tested empirically
fortunately, evidence regarding these hypotheses about harassment thus far, the meta-analytic results support the parts that have been
awaits future research. tested. Among the model’s antecedents, some victim characteris-
tics, especially NA and tenure, are related to harassment, but
organizational characteristics and perpetrator characteristics still
Harassment as a Stressor
need to be studied. The unavailability of studies of perpetrator
Workplace harassment was clearly associated, negatively, with characteristics could be due to the heavy reliance on single-source
victims’ well-being. This supports the idea that harassment is a methods in the research. That is, most of the studies were done by
stressor, having effects similar to other workplace stressors. It is asking for information from victims. This characteristic of the
also interesting that workplace harassment was related to a number research is part of the reason why we formed a model of the
of outcomes even after role conflict and role ambiguity were victim’s view of harassment, but it also is why there is not much
controlled. Role ambiguity and role conflict are stressors that information available on perpetrator characteristics. When survey-
consider the nature of the demands or expectations of people (e.g., ing only victims, it is easy to measure the victim’s NA or PA, for
Cooper & Dewe, 2004), whereas workplace harassment represents example, but it is not possible to measure these characteristics in
a stressor that is a threat to self (Semmer et al., 2005). Both are, the perpetrators.
however, social stressors in the sense that the presumed causal Among harassment’s consequences, victims’ well-being has
stimuli are actions of others in the workplace, highlighting the been a centerpiece of the research, and there is strong support for
importance of the social setting. The incremental variance in this link in the model. Furthermore, the model’s proposal of other
individual outcomes predicted by workplace harassment suggests stressors as partial mediators between harassment and the victim’s
that future research on social aspects of occupational stress should well-being received good support. There is also support for the link
include measures of harassment in addition to measures of more to organizational justice and mixed support for a link to individual
traditional stressors, such as role ambiguity and role conflict. performance outcomes. Links of harassment to turnover intentions,
counterproductive work behaviors, and job performance were sup- victims), and the number of victims and perpetrators present in an
ported but not links to organizational citizenship behaviors and organization. In addition, the model introduces attribution pro-
absenteeism. It should be acknowledged, however, that turnover cesses as an important explanatory variable, and it needs to be
intention and counterproductive work behaviors might be anteced- tested. Future research, therefore, is recommended on the relation-
ents rather than consequences of harassment. In other words, it ships between harassment and these variables. It is especially
seems possible that employees who openly express an interest in important for future research to test the three attributional types
leaving the organization or who engage in counterproductive work (Boxes 8a, 8b, and 8c in the model) as mediating variables be-
behaviors might be target for harassment. The nonexperimental, tween harassment and the outcomes.
cross-sectional data of the studies reviewed did not allow an Another suggestion is the testing of moderator variables.
empirical investigation into this issue, however. There are at least two moderator variables commonly suggested
Overall, therefore, the parts of the model that have been tested by occupational stress theory and research. Social support is
are quite well supported. Obviously however, more research is one (see the meta-analysis by Viswesvaran, Sanchez, & Fisher,
needed on the other parts. Most of the research has examined 1999). Social support might be the antithesis of harassment, if
relationships between harassment and some antecedents and/or it were widespread. That is, if everyone is supportive, then there
some consequences; it has, however, tended to take a “black box” is no harassment. If there is only one supportive person (i.e., the
approach and has ignored potential reasons for the relationships support is not widespread) in the victim’s environment, how-
examined (e.g., antecedent organizational conditions and attribu- ever, then this might be enough to help alleviate the otherwise
tion consequences). harmful effects of harassment. Moderation of stressor–strain
relationships by social support might be most likely to occur
Moderator Analyses when social support matches or is particularly important in
regard to the specific stressor in some way (Cohen & Wills,
In an effort to examine whether the results of the meta-analysis 1985). Social support seems to match the stressor of workplace
were impacted by the overrepresentation of studies using the harassment very well because both are interpersonal in nature
ICAWS (Spector & Jex, 1998) to measure harassment, we con- and because support can show respect to the victim that might
ducted moderator analyses to examine whether the ICAWS counteract the threat to esteem that comes from harassment.
yielded different effect sizes than did non-ICAWS measures. For Future research based on the matching hypothesis and examin-
many variables, there were no practical differences. When practi- ing potential interactions between workplace harassment as a
cal differences did exist, effect sizes were generally greater for the stressor and social support is needed.
ICAWS than for non-ICAWS measures. Even when practical A second theoretical moderator common in occupational stress
differences did exist, however, the direction of the relationships research is job control, which is a major element in demand–
between harassment and the other variables were the same for both control theory of occupational stress (e.g., see the review by Van
the ICAWS and non-ICAWS measures. Thus, the results obtained Der Doef & Maes, 1999). Control over a situation tends to make
in the current meta-analysis concerning the direction of the rela- otherwise stressful situations less distressing, but exactly what type
tionships between harassment and its predictors and consequences of control would be most effective in a harassing situation is
were not influenced by the overrepresentation of ICAWS studies. undetermined. It is unfortunate that empirical research on harass-
In other words, the overrepresentation of ICAWS studies did not ment has not examined theoretical moderators such as these, and
impact the general conclusions provided by the current this therefore remains a goal for future research.
meta-analysis. Future research should also address the attributional pro-
We should note, however, that the differences in effect sizes cesses that occur when one is harassed. In addition to testing the
between ICAWS and non-ICAWS studies is potentially the result model with the type of harassment proposed in the current
of artifacts rather than being indicative of a substantive moderating article, research could examine the role of attributions in (a)
effect. More specifically, second-order sampling error, which is a vicarious workplace harassment and (b) sexual harassment.
problem among small k meta-analyses, occurs when the results of Individuals could report the presence of vicarious harassment
a meta-analysis are influenced by which studies happen to be when they perceive that their coworkers are victims. Vicarious
available to the researcher (Hunter & Schmidt, 2004). This is harassment is likely to produce negative consequences even in
consistent with the fact that most of the moderator analyses that employees who are not victims because such individuals could
appeared to detect practically significant differences in effect sizes empathize with the victim, make cognitive judgments about the
between the two types of harassment measures were generally for situation (e.g., justice judgments), and have a reaction about a
variables with a small number of studies (e.g., k of 10 or less) using workplace in which such activities occur. In some instances,
either the ICAWS or non-ICAWS measures. they could of course fear that they too will become victims—
perhaps especially in situations in which the victims resembled
Future Research themselves in some way (e.g., hierarchical position or job type).
It could thus be expected that vicarious harassment has conse-
The stress theme suggests additional topics for future empirical quences that are similar to but milder than other forms of
work. Some specific variables in the model still need empirical harassment. This needs to be tested in future research, however.
research to link them to workplace harassment. These include the Attributional processes are also likely to be relevant to sexual
organizational variables of culture and climate, the nature of hu- harassment. The victim could blame herself (assuming women
man resource systems that might be responsible for the existence are the more common target of sexual harassment), resulting in
of people with characteristics of likely perpetrators (or likely self-deprecating attitudes and perhaps in behaviors such as
leaving the workplace. Alternatively, she could blame the per- be a coworker who is another union member. This is an area for
petrator and develop negative attitudes and behaviors mainly future research.
toward him (assuming men are more commonly the perpetra-
tors). If she attributes blame primarily to the organization, Implications and Limitations
however, sexual harassment complaints and lawsuits would be
more likely to follow. The findings of the current meta-analysis have important
practical implications. First, it appears that organizations might
be able to combat workplace harassment by eliminating or
Individual Performance and Satisfaction reducing other workplace stressors. Although the reduction of
The current meta-analysis also found a relationship with workplace stressors is important for its own sake and a number
some individual performance outcomes, such as counterproduc- of interventions have been developed to reduce workplace
tive work behaviors and turnover intention. Of interest, harass- stressors, employers seldom actually use such interventions for
ment was not strongly related to job performance. Job perfor- stress reduction (Beehr & O’Driscoll, 2002). Regarding role
mance is multidimensional, and it is possible that harassment is stressors, the type of stressors investigated in the regression
differentially related to different performance dimensions. Be- analysis in the present study, such interventions include role
ing harassed by a supervisor, for instance, could increase em- clarification and reducing competing demands on people’s time,
ployees’ quantitative performance while inhibiting their quali- but interventions aimed at other types of work stressors include
tative performance. Studies included in the current meta- increasing social support in the workplace (almost the opposite
analysis, however, reported only the relationships between of workplace harassment), participation in decision making, and
harassment and global performance. Furthermore, the relative some structural changes (Beehr & O’Driscoll, 2002). Social
small number of studies examining the relationship between support as a stress reducer seems so clearly related to harass-
harassment and job performance (k ⫽ 5) precludes any moder- ment (negatively) that it might be especially effective, although
ator analysis for these relationships. research is needed to verify this.
Similar to job performance, job satisfaction is multidimensional, The relationship between NA and workplace harassment also
that is, there are facets of satisfaction. The relationship between has important practical implications for preventing victimiza-
workplace harassment and facet-specific job satisfactions has been tion. Organizations might closely monitor those high in NA to
examined very little, but some relationship was evident (e.g., prevent them from becoming victims of harassment, for in-
Cortina et al., 2001; Keashly, Trott, & MacLean, 1994). Future stance. Unfortunately, little research has been conducted on the
empirical study of facet satisfactions is warranted. The “social” characteristics of potential perpetrators of harassment. Once it
facets (e.g., satisfaction with supervisors and coworkers) seem is clearly understood who is likely to become a perpetrator,
especially likely foci for such research because other people are organizations could use selection tests to screen-out potentially
the source of workplace harassment. Particularly, harassment per- aggressive employees or organizations could closely monitor
petrated by coworkers is likely to be related to satisfaction with individuals who are prone to become harassers. This approach,
coworkers and harassment perpetrated by supervisors is likely to however, awaits further research to identify perpetrator
be related to satisfaction with supervision. All harassment, regard- characteristics.
less of the type of perpetrator, is likely to be related to satisfaction It is important to note certain limitations of the current
with supervision because most employees are likely to consider it meta-analysis. First, the studies included in the analyses used
the responsibility of management to prevent their employees from all self-report measures. Thus, common-method variance might
being victimized. have artificially inflated some of the correlations that we ob-
Because of the negative outcomes associated with workplace tained. Second, the fact that all of the studies also used cross-
harassment, it is unfortunate that state and federal governments sectional and nonexperimental data makes it impossible to draw
as well as employers do not generally regard harassment as a strong conclusions concerning the causal relationships between
greater priority. Legislation, guidelines for the promotion of harassment and the other variables included in the study. Fi-
nonharassing workplaces, and budgets for research into the nally, all the studies examined workplace harassment only from
topic appear warranted. Organizations also have something to the perspective of the victim. Each of these methodological
gain from giving more of their attention to harassment. Some of characteristics of the previous research has obvious implica-
the consequences of harassment are likely to impact the effec- tions for future research needs.
tiveness and functioning of the employing organization, includ-
ing the potential effects of harassment on employees’ job sat- References
isfaction, organizational commitment, perceptions of
References marked with an asterisk indicate studies included in the
organizational justice, and employee withdrawal. Research on meta-analysis.
organizational efforts aimed at the reduction of harassment is
needed. Workplace harassment is also an issue in which labor *Agervold, M., & Mikkelsen, E. G. (2004). Relationships between bully-
ing, psychosocial work environment and individual stress reactions.
unions should take a keen interest, considering their historical
Work and Stress, 18, 336 –351.
concern for worker well-being. Unions have traditionally *Allen, S. J. (1993). The moderating effects of self-esteem: A closer look at
played an important role in protecting their members from the relationship between work stress and employee health. Unpublished
abuses by the employer or the employers’ representatives (e.g., master’s thesis, Central Michigan University, Mount Pleasant.
supervisors and managers), but they may or may not see their *Allen, S. J. (1995). An examination of the relationship between two types
protective role as clearly in cases where the perpetrator might of occupational stressors: Chronic stressors and traumatic events. Un-
published doctoral dissertation, Central Michigan University, Mount Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering
Pleasant. hypothesis. Psychological Bulletin, 98, 310 –357.
Andersson, L. M., & Pearson, C. M. (1999). Tit for tat? The spiraling effect Collins, V. A. (1999). A meta-analysis of burnout and occupational stress.
of incivility in the workplace. Academy of Management Review, 24, Unpublished doctoral dissertation, University of North Texas, Denton.
452– 471. Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O. L. H., & Ng,
Aquino, K. (2000). Structural and individual determinants of workplace K. Y. (2001). Justice at the millennium: A meta-analytic review of 25
victimization: The effects of hierarchical status and conflict management years of organizational justice research. Journal of Applied Psychology,
style. Journal of Management, 26, 171–193. 86, 425– 445.
*Aquino, K., Grover, S. L., Bradfield, M., & Allen, D. G. (1999). The Cooper, C. L., & Dewe, P. (2004). Stress: A brief history. Malden, MA:
effects of negative affectivity, hierarchical status, and self-determination Blackwell.
on workplace victimization. Academy of Management Journal, 42, 260 – *Cortina, L. M., Magley, V. J., Williams, J. H., & Langhout, R. D. (2001).
272. Incivility in the workplace: Incidence and impact. Journal of Occupa-
Aquino, K., & Lamertz, K. (2004). A relational model of workplace tional Health Psychology, 6, 64 – 80.
victimization: Social roles and patterns of victimization in dyadic rela- Crawford, J. S. (2000). Comparative analysis of chronic versus acute
tionships. Journal of Applied Psychology, 89, 1023–1034. stressors and their influence on distress. Unpublished doctoral disserta-
Ashforth, B. E. (1994). Petty tyranny in organizations. Human Relations, tion, University of North Texas, Denton.
47, 755–778. *Duffy, M. K., & Ferrier, W. J. (2003). Birds of a feather. . .? How
*Ashforth, B. E. (1997). Petty tyranny in organizations: A preliminary supervisor–subordinate dissimilarity moderates the influence of super-
examination of antecedents and consequences. Canadian Journal of visor behaviors on workplace attitudes. Group and Organization Man-
Administrative Sciences, 14, 126 –140. agement, 28, 217–248.
Baldwin, D. C., Daugherty, S. R., & Eckenfels, E. J. (1991). Student *Duffy, M. K., Ganster, D. C., & Pagon, M. (2002). Social undermining in
perceptions of mistreatment and harassment during medical school: A the workplace. Academy of Management Journal, 45, 331–351.
survey of ten United States schools. Western Journal of Medicine, 155, Einarsen, S. (2000). Harassment and bullying at work: A review of the
140 –145. Scandinavian approach. Aggression and Violent Behavior, 5, 379 – 401.
*Barling, J., Rogers, A. G., & Kelloway, E. K. (2001). Behind closed *Einarsen, S., Matthiesen, S. B., & Skogstad, A. (1998). Bullying, burnout
doors: In-home workers’ experience of sexual harassment and work- and well-being among assistant nurses. Journal of Occupational Health
place violence. Journal of Occupational Health Psychology, 6, 255–269. and Safety—Australia and New Zealand, 14, 563–568.
Beehr, T. A., & Glazer, S. (2005). Organizational role stress. In J. Barling, Einarsen, S., & Mikkelsen, E. G. (2003). Individual effects of exposure to
E. K. Kelloway, & M. R. Frone (Eds.), Handbook of work stress (pp. bullying at work. In S. Einarsen, H. Hoel, D. Zapf, & C. L. Cooper
7–33). Thousand Oaks, CA: Sage. (Eds.), Bullying and emotional abuse in the workplace: International
Beehr, T. A., & O’Driscoll, M. P. (2002). Organizationally targeted inter- perspectives in research and practice (pp. 127–144). London: Taylor &
ventions aimed at reducing workplace stress. In J. C. Thomas & M. Francis.
Hersen (Eds.), Handbook of mental health in the workplace (pp. 103– *Einarsen, S., & Raknes, B. I. (1997). Harassment in the workplace and the
119). Thousand Oaks, CA: Sage. victimization of men. Violence and Victims, 12, 247–263.
Bies, R. J., & Tripp, T. M. (2005). The study of revenge in the workplace: *Einarsen, S., Raknes, B. I., & Matthiesen, S. B. (1994). Bullying and
Conceptual, ideological, and empirical issues. In S. Fox & P. E. Spector harassment at work and their relationships to work environment quality:
(Eds.), Counterproductive work behavior: Investigations and actors and An exploratory study. European Work and Organizational Psychologist,
targets (pp. 65– 81). Washington, DC: American Psychological Associ- 4, 381– 401.
ation. *Einarsen, S., & Skogstad, A. (1996). Bullying at work: Epidemiological
*Bjorkqvist, K., Osterman, K., & Hjelt-Back, M. (1994). Aggression findings in public and private organizations. European Journal of Work
among university employees. Aggressive Behavior, 20, 173–184. and Organizational Psychology, 5, 185–201.
Bjorkqvist, K., Osterman, K., & Lagerspetz, K. M. J. (1994). Sex differ- *Factor, M. H. (1998). Sense of humor and social support as moderators
ences in covert aggression among adults. Aggressive Behavior, 20, of the occupational stressor–strain relationship: An exploratory field
27–33. investigation. Unpublished doctoral dissertation, University of South
*Bowling, N. A., Beehr, T. A., & Swader, W. M. (2004, April). Anteced- Florida, Tampa.
ents to giving and receiving social support at work. Paper presented at *Fortunato, V. J. (2004). A comparison of the construct validity of three
the annual meeting of the Society for Industrial and Organizational measures of negative affectivity. Educational and Psychological Mea-
Psychology, Chicago, IL. surement, 64, 271–289.
*Bowling, N. A., Watson, C. P., & Beehr, T. A. (2004, April). Situational *Fox, S., Spector, P. E., Goh, A., & Bruursema, K. (2003, April). Volun-
and dispositional factors in job and life satisfaction. Paper presented at tary work behavior: Exploring parallels between counterproductive
the annual meeting of the Society for Industrial and Organizational work behavior and organizational citizenship behavior. Paper presented
Psychology, Chicago, IL. at the annual meeting of the Society for Industrial and Organizational
Budd, J. W., Arvey, R. D., & Lawless, P. (1996). Correlates and conse- Psychology, Orlando, FL.
quences of workplace violence. Journal of Occupational Health Psy- *Fox, S., Spector, P. E., & Miles, D. (2001). Counterproductive work
chology, 1, 197–210. behavior (CWB) in response to job stressors and organizational justice:
*Chen, P. Y., & Spector, P. E. (1991). Negative affectivity as the under- Some mediator and moderator tests for autonomy and emotions. Journal
lying cause of correlations between stressors and strains. Journal of of Vocational Behavior, 59, 291–309.
Applied Psychology, 76, 398 – 407. *Frone, M. R. (1998a). Predictors of work injuries among employed
*Chen, P. Y., & Spector, P. E. (1992). Relationships of work stressors with adolescents. Journal of Applied Psychology, 83, 565–576.
aggression, withdrawal, theft and substance use: An exploratory study. *Frone, M. R. (1998b, April). Predictors of workplace deviance among
Journal of Occupational and Organizational Psychology, 65, 177–184. employed adolescents. Paper presented at the annual meeting of the
Cialdini, R. B. (2001). Influence: Science and practice. Needham Heights, Society for Industrial and Organizational Psychology, Dallas, TX.
MA: Allyn & Bacon. *Frone, M. R. (2000). Interpersonal conflict at work and psychological
outcomes: Testing a model among young workers. Journal of Occupa- of emotional abuse in the workplace: Results of a statewide survey.
tional Health Psychology, 5, 246 –255. Paper presented at the annual meeting of the Academy of Management,
Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader– Toronto, Ontario, Canada.
member exchange theory: Correlates and construct issues. Journal of *Keashly, L., Trott, V., & MacLean, L. M. (1994). Abusive behavior in the
Applied Psychology, 86, 827– 844. workplace: A preliminary investigation. Violence and Victims, 9, 341–
*Glomb, T. M., & Liao, H. (2003). Interpersonal aggression in work 357.
groups: Social influence, reciprocal, and individual effects. Academy of Kelley, H. H. (1973). The process of causal attribution. American Psychol-
Management Journal, 46, 486 – 496. ogist, 28,107–128.
Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. *Kruse, B. G. (2003). The role of causal attributions in the relationship
American Sociological Review, 25, 161–178. between workplace stressors and social support. Unpublished doctoral
*Grandey, A. A., Dickter, D. N., & Sin, H. P. (2004). The customer is not dissertation, Central Michigan University, Mount Pleasant.
always right: Customer aggression and emotion regulation of service Lamertz, K., & Aquino, K. (2004). Social power, social status and percep-
employees. Journal of Organizational Behavior, 25, 397– 418. tual similarity of workplace victimization: A social network analysis of
Greenberg, L., & Barling, J. (1999). Predicting employee aggression stratification. Human Relations, 57, 795– 822.
against coworkers, subordinates and supervisors: The roles of person Lapierre, L. M., Spector, P. E., & Leck, J. D. (2005). Sexual versus
behaviors and perceived workplace factors. Journal of Organizational nonsexual workplace aggression and victims’ overall job satisfaction: A
Behavior, 20, 897–913. meta-analysis. Journal of Occupational Health Psychology, 10, 155–
*Heinisch, D. A. (1995). A comparison of self-report vs. observer mea- 169.
sures of negative affectivity in the study of work-related stress. Unpub- *LeBlanc, M. M., & Kelloway, E. K. (2002). Predictors and outcomes of
lished doctoral dissertation, Central Michigan University, Mount Pleas- workplace violence and aggression. Journal of Applied Psychology, 87,
ant. 444 – 453.
*Heinisch, D. A., & Jex, S. M. (1997). Negative affectivity and gender as *Lee, V. B. (2003). The impact of different sources of interpersonal
moderators of the relationship between work-related stressors and de- conflict at work on the target of counterproductive work behavior.
pressed mood at work. Work and Stress, 11, 46 –57. Unpublished master’s thesis, University of South Florida, Tampa.
Hoel, H., Einarsen, S., & Cooper, C. L. (2003). Organizational effects of Leymann, H. (1996). The content and development of mobbing at work.
bullying. In S. Einarsen, H. Hoel, D. Zapf, & C. L. Cooper (Eds.), European Journal of Work and Organizational Psychology, 5, 165–184.
Bullying and emotional abuse in the workplace: International perspec- *Liu, C. (2003). A comparison of job stressors and job strains among
tives in research and practice (pp. 145–161). London: Taylor & Francis. employees holding comparable jobs in Western and Eastern societies.
*Hoel, H., Faragher, B., & Cooper, C. L. (2004). Bullying is detrimental to Unpublished doctoral dissertation, University of South Florida, Tampa.
health, but all bullying behaviours are not necessarily equally damaging. *Lubbers, R. W. (2003). Self-efficacy and affective well-being among
British Journal of Guidance and Counselling, 32, 367–387. young workers: Examining job quality as an antecedent of employee
Hoel, H., Rayner, C., & Cooper, C. L. (1999). Workplace bullying. health and performance outcomes. Unpublished doctoral dissertation,
International Review of Industrial and Organizational Psychology, 14, University of Toronto, Toronto, Ontario, Canada.
195–230. *Marrs, M. E. M. (1999). Antecedents and outcomes of verbal aggression
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Cor- in the workplace. Unpublished doctoral dissertation, University of Mis-
recting error and bias in research findings. Thousand Oaks, CA: Sage. souri, Columbia.
Jackson, S. E., & Schuler, R. S. (1985). A meta-analysis and conceptual McCrae, R. R., & Costa, P. T. (1987). Validation of the five-factor model
critique of research on role ambiguity and role conflict in work settings. of personality across instruments and observers. Journal of Personality
Organizational Behavior and Human Decision Processes, 36, 16 –78. and Social Psychology, 52, 81–90.
*Jayaratne, S., Vinokur-Kaplan, D., Nagda, B. A., & Chess, W. A. (1996). *Menon, S., Narayanan, L., & Spector, P. E. (1996). Time urgency and its
A national study on violence and harassment of social workers by relation to occupational stressors and health outcomes for health care
clients. Journal of Applied Social Sciences, 1, 1–14. professionals. In C. D. Spielberger, I. G. Sarason, J. M. T. Brebner, E.
*Jex, S. M., Adams, G. A., Elacqua, T. C., & Lux, D. J. (1997). A Greenglass, P. Laungani, & A. M. O’Roark (Eds.), Stress and emotions:
comparison of incident-based and scale measures of work stressors. Anxiety, anger, and curiosity (Vol. 16, pp. 127–142). London: Taylor &
Work and Stress, 11, 229 –238. Francis.
*Jex, S. M., Beehr, T. A., & Roberts, C. K. (1992). The meaning of *Mikkelsen, E. G., & Einarsen, S. (2001). Bullying in Danish work-life:
occupational stress items to survey respondents. Journal of Applied Prevalence and health correlates. European Journal of Work and Orga-
Psychology, 77, 623– 628. nizational Psychology, 10, 393– 413.
*Jex, S. M., & Spector, P. E. (1996). The impact of negative affectivity on *Mikkelsen, E. G., & Einarsen, S. (2002). Relationships between exposure
stressor–strain relationships: A replication and extension. Work and to bullying at work and psychological and psychosomatic health com-
Stress, 10, 36 – 45. plaints: The role of state negative affectivity and generalized self-
*Jex, S. M., & Thomas, J. L. (2003). Relations between stressors and group efficacy. Scandinavian Journal of Psychology, 43, 397– 405.
perceptions: Main and mediating effects. Work and Stress, 17, 158 –169. *Miles, D. E., Borman, W. E., Spector, P. E., & Fox, S. (2002). Building
*Johnson, A. L. (2003). Dispositional and situational predictors of cyni- an integrative model of extra role work behaviors: A comparison of
cism about organizational change. Unpublished master’s thesis, Central counterproductive work behavior with organizational citizenship behav-
Michigan University, Mount Pleasant. ior. International Journal of Selection and Assessment, 10, 51–57.
*Kaukiainen, A., Salmivalli, C., Bjorkqvist, K., Osterman, K., Lahtinen, *Miner-Rubino, K. N. (2004). Beyond targets: Vicarious exposure to
A., Kostamo, A., & Lagerspetz, K. (2001). Overt and covert aggression hostility toward women in the workplace. Unpublished doctoral disser-
in work settings in relation to the subjective well-being of employees. tation, University of Michigan, Ann Arbor.
Aggressive Behavior, 27, 360 –371. *Moberg, P. J., Ritter, B., & Fischbein, R. (2002, August). Predicting
*Keashly, L., Hunter, S., & Harvey, S. (1997). Abusive interaction and role abusive managerial behavior: Antecedents of supervisory job perfor-
state stressors: Relative impact on student residence assistant stress and mance. Paper presented at the annual meeting of the Academy of
work attitudes. Work and Stress, 11, 175–185. Management, Denver, CO.
*Keashly, L., & Jagatic, K. (2000, August). The nature, extent, and impact Neuman, J. H., & Baron, R. A. (1997). Aggression in the workplace. In
R. A. Giacalone & J. Greenberg (Eds.), Antisocial behavior in organi- (1990). A pilot study of medical student “abuse.” Journal of the Amer-
zations (pp. 37– 67). Thousand Oaks, CA: Sage. ican Medical Association, 263, 533–537.
*Neuman, J. H., & Keashly, L. (2003, April). Workplace bullying: Persis- Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and non-
tent patterns of workplace aggression. In J. Greenberg & M. E. Roberge experimental studies: New procedures and recommendations. Psycho-
(Chairs), Vital but neglected topics in workplace deviance research. logical Methods, 7, 422– 445.
Symposium conducted at the annual meeting of the Society for Industrial Silver, H. K., & Glicken, A. D. (1990). Medical student abuse. Journal of
and Organizational Psychology, Orlando, FL. the American Medical Association, 263, 527–532.
*Niedl, K. (1996). Mobbing and well-being: Economic and personnel Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The role
development implications. European Journal of Work and Organiza- of distributive, procedural, and interaction justice. Journal of Applied
tional Psychology, 5, 239 –249. Psychology, 82, 434 – 443.
*Papinchock, J. M. (1992). An investigation of the relations of shiftwork, Skjorshammer, M., & Hofoss, D. (1999). Physician in conflict: A survey
sleep, the use of chemical substances, and stress. Unpublished doctoral study of individual and work-related characteristics. Scandinavian Jour-
dissertation, University of South Florida, Tampa. nal of Caring Science, 13, 211–216.
*Penney, L. M. (2003). Workplace incivility and counterproductive work- *Spector, P. E. (1987). Interactive effects of perceived control and job
place behavior (CWB): What is the relationship and does personality stressors on affective reactions and health outcomes for clerical workers.
play a role? Unpublished doctoral dissertation, University of South Work and Stress, 1, 155–162.
Florida, Tampa. *Spector, P. E. (1991). Confirmatory test of a turnover model utilizing
*Potter, P. T., Smith, B. W., Strobel, K. R., & Zautra, A. J. (2002). multiple data sources. Human Performance, 4, 221–230.
Interpersonal workplace stressors and well-being: A multi-wave study of *Spector, P. E., Dwyer, D. J., & Jex, S. M. (1988). Relation of job stressors
employees with and without arthritis. Journal of Applied Psychology, 87, to affective, health, and performance outcomes: A comparison of mul-
789 –796. tiple data sources. Journal of Applied Psychology, 73, 11–19.
*Quine, L. (2001). Workplace bullying in nurses. Journal of Health Psy- Spector, P. E., & Jex, S. M. (1998). Development of four self-report
chology, 6, 73– 84. measures of job stressors and strain: Interpersonal Conflict at Work
*Rainey, D. W. (1995). Stress, burnout, and intention to terminate among Scale, Organizational Constraints Scale, Quantitative Workload Inven-
tory, and Physical Symptoms Inventory. Journal of Occupational Health
umpires. Journal of Sport Behavior, 18, 312–323.
Psychology, 3, 356 –366.
*Rainey, D. W. (1999). Sources of stress, burnout, and intention to termi-
*Spector, P. E., & O’Connell, B. J. (1994). The contribution of personality
nate among basketball referees. Journal of Sport Behavior, 22, 578 –590.
traits, negative affectivity, locus of control and Type A to the subsequent
Rayner, C. (1998). Workplace bullying: Do something! Journal of Occu-
reports of job stressors and job strains. Journal of Occupational and
pational Health Safety, 14, 581–585.
Organizational Psychology, 67, 1–11.
Rayner, C., & Hoel, H. (1997). A summary review of literature relating to
*Tepper, B. J. (2000). Consequences of abusive supervision. Academy of
workplace bullying. Journal of Community and Applied Social Psychol-
Management Journal, 43, 178 –190.
ogy, 7, 181–191.
*Tepper, B. J., Duffy, M. K., Hoobler, J., & Ensley, M. D. (2006).
Richman, J. A., Flaherty, J. A., Rospenda, K. M., & Christensen, M. L.
Moderating effects of abusive supervision on relationships between
(1992). Mental health consequences and correlates of reported medical
coworkers’ organizational citizenship behavior and fellow employees’
student abuse. Journal of the American Medical Association, 267, 692–
attitudes. Unpublished manuscript.
694.
*Tepper, B. J., Duffy, M. K., & Shaw, J. D. (2001). Personality moderators
Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace
of the relationship between abusive supervision and subordinates’ resis-
behaviors: A multidimensional scaling study. Academy of Management
tance. Journal of Applied Psychology, 86, 974 –983.
Journal, 38, 555–572. Van Der Doef, M., & Maes, S. (1999). The job demand– control (–support)
*Rogers, K. A., & Kelloway, E. K. (1997). Violence at work: Personal and model and psychological well-being: A review of 20 years of empirical
organizational outcomes. Journal of Occupational Health Psychology, research. Work and Stress, 13, 87–114.
2, 63–71. *Vartia, M. (2001). Consequences of workplace bullying with respect to
*Rospenda, K. M. (1999). Sexual and non-sexual harassment as interper- the well-being of its targets and the observers of bullying. Scandinavian
sonal conflict stressors in the workplace: Effects on job, mental health, Journal of Work, Environment and Health, 27, 63– 69.
and physical health outcomes. Unpublished doctoral dissertation, De- *Vartia, M., & Hyyti, J. (2002). Gender differences in workplace bullying
paul University, Chicago. among prison officers. European Journal of Work and Organizational
Rospenda, K. M., Richman, J. A., Wislar, J. S., & Flaherty, J. A. (2000). Psychology, 11, 113–126.
Chronicity of sexual harassment and generalized work-place abuse: Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psy-
Effects on drinking outcomes. Addiction, 95, 1805–1820. chometric meta-analysis with structural equation modeling. Personnel
*Schat, A. C. H., & Kelloway, E. K. (2000). Effects of perceived control Psychology, 48, 865– 885.
on the outcomes of workplace aggression and violence. Journal of Viswesvaran, C., Sanchez, J. I., & Fisher, J. (1999). The role of social
Occupational Health Psychology, 5, 386 – 402. support in the process of work stress: A meta-analysis. Journal of
*Schat, A. C. H., & Kelloway, E. K. (2003). Reducing the adverse Vocational Behavior, 54, 314 –334.
consequences of workplace aggression and violence: The buffering *Winstanley, S., & Whittington, R. (2002). Anxiety, burnout and coping
effects of organizational support. Journal of Occupational Health Psy- styles in general hospital staff exposed to workplace aggression: A
chology, 8, 110 –122. cyclical model of burnout and vulnerability to aggression. Work and
Semmer, N. K., McGrath, J. E., & Beehr, T. A. (2005). Conceptual issues Stress, 16, 302–315.
in research on stress and health. In C. L. Cooper (Ed.), Handbook of Wolf, T. M., Randall, H. M., Von Almen, K., & Tynes, L. L. (1991).
stress medicine and health (2nd ed., pp. 1– 43). London: CRC Press. Perceived mistreatment and attitude change by graduating medical stu-
Shaw, J. C., Wild, E., & Colquitt, J. A. (2003). To justify or excuse?: A dents: A retrospective study. Medical Education, 25, 182–190.
meta-analytic review of the effects of explanations. Journal of Applied Zapf, D., Einarsen, S., Hoel, H., & Vartia, M. (2003). Empirical
Psychology, 88, 444 – 458. findings on bullying. In S. Einarsen, H. Hoel, D. Zapf, & C. L.
Sheehan, H., Sheehan, D. V., White, K., Leibowitz, A., & Baldwin, D. C. Cooper (Eds.), Bullying and emotional abuse in the workplace:
International perspectives in research and practice (pp. 103–126). Zuckerman, M. (1979) Attribution of success and failure revisited, or: The
London: Taylor & Francis. motivational bias is alive and well in attribution theory. Journal of
*Zapf, D., Knorz, C., & Kulla, M. (1996). On the relationship between Personality, 47, 245–287.
mobbing factors, and job content, social work environment, and health
outcomes. European Journal of Work and Organizational Psychology,
5, 215–237.
*Zellars, K. L., Tepper, B. J., & Duffy, M. K. (2002). Abusive supervision Received May 24, 2004
and subordinates’ organizational citizenship behavior. Journal of Ap- Revision received October 26, 2005
plied Psychology, 87, 1068 –1076. Accepted November 4, 2005 䡲
View publication stats

Workplace Harassment From The Victim's Perspective: A Theoretical Model and Meta-Analysis

Uploaded by

Copyright:

Available Formats

Workplace Harassment From The Victim's Perspective: A Theoretical Model and Meta-Analysis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Workplace Harassment From The Victim's Perspective: A Theoretical Model and Meta-Analysis

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Workplace Harassment from the Victim's Perspective: A Theoretical Model

Article in Journal of Applied Psychology · October 2006

Research and Theory on Workplace Aggression View project

The user has requested enhancement of the downloaded file.

Workplace Harassment From the Victim’s Perspective: A Theoretical

(1) Organizational Culture/Climate

(2) Organizational Human Resource

(3) Perpetrator (4) Victim

(5) Workplace Harassment (6) Role

(7) Number (8) Process of

(11a) (10) Attitudes

% variance 80% credibility

Role conflict 16 5,429 .35 .44 .08 37.39 .34, .55

% variance 80% credibility

Generic strain 27 17,663 .29 .35 .08 23.00 .24, .46

% variance 80% credibility

Organizational justice 8 2,061 ⫺.29 ⫺.35 .13 23.22 ⫺.51,⫺.18

Harassment Role ambiguity 6,996 .14** .21**

Discussion predispose employees to engage in behaviors that provoke mis-

rassed by supervisors, agreeable and conscientious subordinates The Model

View publication stats

You might also like

Harassment Role ambiguity 6,996 .14 .21