A Model of Work Frustration Aggression
Uploaded by
Esraa TalaatA Model of Work Frustration Aggression
Uploaded by
Esraa TalaatJournal of Organizational Behavior
J. Organiz. Behav. 20, 915±931 (1999)
A model of work frustration±aggression
SUZY FOX* AND PAUL E. SPECTOR
Department of Psychology, University of South Florida, U.S.A.
Summary The current study was designed to investigate the situational, dispositional, and aective
antecedents of counterproductive work behaviors. A model based on the organizational
frustration±aggression work of Spector and colleagues was tested using structural equa-
tion modeling and zero-order correlational analysis. As expected, a positive relationship
was found between employees' experience of situational constraints (events frustrating
their achievement of organizational and personal goals) and counterproductive beha-
vioral responses to frustration (personal and organizational aggression), mediated by
aective reactions to frustration. In addition, personality (trait anger and trait anxiety),
control beliefs (Work Locus of Control), and estimation of likelihood of punishment
were strongly associated with aective and behavioral responses. In particular, strong
direct relationships were found between aective response variables and anxiety and
locus of control, while direct relationships were found between behavioral response
variables and anger and punishment. Finally, dierentiated relationships between two
facets of trait anger (angry temperament and angry reaction) and four categories of
counterproductive behaviors (serious and minor deviance directed at organizational and
personal targets) were explored. Copyright # 1999 John Wiley & Sons, Ltd.
Introduction
Success in the postmodern, post-industrial organization is in part a function of employees'
behavioral role choices. The employee may choose a role that goes beyond the performance
requirements de®ned by the job role itself, or conversely may choose a role that covertly or
overtly harms the organization. Thus one broad stream of organizational and management
research looks at the need of the competitive organization for employees whose performance
exceeds the demands and expectations of their positions. Such role choices are variously studied
as pro-role (McLean Parks and Kidder, 1994), prosocial (Brief and Motowidlo, 1986), and
organizational citizenship (Organ, 1990) behavior. A second area of research looks at the
vulnerability of organizations to employees' role choices that may harm the organization, ranging
from mere compliance to hostility, organizational aggression, or sabotage. This set of role
* Correspondence to: Suzy Fox, Department of Psychology, University of South Florida, 4202 E. Fowler BEH 339,
Tampa, FL 33620±8200; U.S.A. E-mail: [email protected]
This article was based, in part, on research carried out for Suzy Fox's master of psychology thesis under the direction of
Paul E. Spector. Portions of this research were presented in August 1997 as a poster at the 105th Annual Convention of
the American Psychological Association, Chicago, Illinois.
CCC 0894±3796/99/060915±17$17.50
Copyright # 1999 John Wiley & Sons, Ltd. Accepted 26 August 1997
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
916 S. FOX AND P. E. SPECTOR
choices is being studied as anti-role (McLean Parks and Kidder, 1994), antisocial (Hogan and
Hogan, 1989), counterproductive (Storms and Spector, 1987), deviant (Robinson and Bennett,
1995) or maladaptive (Perlow and Latham, 1993) work behavior.
Cognitive versus aective components
of attitude formation
What many of the studies in this behavioral role choice area of research have in common is a
fundamental focus on the cognitive basis of work-related attitude formation, such as percep-
tions of organizational fairness, equity theory, and the Fishbein and Ajzen (1975) Theory of
Reasoned Action. In the aforementioned studies, the aective component of attitude formation
or behavioral response is frequently mentioned, but set aside in the obligatory suggestions for
further research. This near-exclusive focus on the cognitive component misses a large part of the
explanation of attitudinal and behavioral variance.
Frustration±aggression and counterproductive
organizational behavior
Recently, however, there has been some resurgence of interest in the aective component of
attitude formation and behavioral response. One model of the aective basis of organizational
aggression has been developed by Spector and colleagues (Chen and Spector, 1992; Spector,
1975, 1978, 1997; Storms and Spector, 1987). This model has its roots in the classic Dollard±
Miller frustration±aggression theory (Dollard, Doob, Miller, Mowrer and Sears, 1939). It
focuses on the interplay of aective and behavioral responses to certain types of work situations.
The Dollard±Miller model views aggression as a consequence of frustration. A frustration
occurs when an instigated goal-response (or predicted behavioral sequence) is interrupted or
interdicted. It is possible that the individual may ®nd a substitute response for the prevented goal-
response; however, if that does not occur, the individual may respond with some level (overt or
covert, externally or internally directed) of aggression (Dollard et al., 1939). The form this
aggressive response takes will be strongly in¯uenced by the individual's perception of the
likelihood of being punished. According to Dollard et al. (1939) `the inhibition of any act of
aggression varies directly with the strength of the punishment anticipated for the expression of
that act' (p. 37). Thus overt expressions of aggression through counterproductive organizational
behavior would be expected to be related to the perception that one could `get away with it'
without being caught or punished.
Subsequent research on this frustration±aggression sequence highlights the mediation of
emotional reaction. In particular, Berkowitz critiqued the original frustration±aggression hypo-
thesis for neglecting the mediating role of the arousal of negative aect on the relationship
between frustrations and ®ght-or-¯ight behavior (Spielberger, Reheiser and Sydeman, 1995). The
two important aspects of this reaction are that the emotion is aversive, and that it results in
increased physiological arousal (Spector, 1978). Another criticism of Dollard et al. is that their
approach was too mechanistic, ignoring cognitive and dispositional processes. A more compre-
hensive model would consider the eects of belief-based variables such as locus of control, as well
as personality dispositions.
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 917
Spector and colleagues extended this theory to the phenomenon of organizational aggression.
Their models examine the sequence of frustrated events, emotional reactions to frustration, and
behavioral reactions, in organizational work. Frustrated events may be understood as situational
constraints in the immediate work situation that block individuals from achieving valued work
goals or attaining eective performance (Peters and O'Connor, 1980). Emotional reactions to
frustrated events can be de®ned both as perceived frustration (Storms and Spector, 1987) and the
resulting negative emotional state, consisting of aversion and arousal (Spector, 1978). This
emotional reaction may be considered a subdomain of the much-studied job stress area. Aective
reactions include job dissatisfaction, feelings of stress, feelings of frustration, anxiety and anger
(Chen and Spector, 1992).
Behavioral reactions to organizational frustration include eects on job performance, absent-
eeism, turnover, organizational aggression, and the interpersonal aggression. To the extent that
these behaviors interfere with the organization's task performance, climate, or eectiveness, they
may tangibly damage the organization. As such they may be thought of as counterproductive,
anti-role, antisocial, maladaptive or deviant behaviors. These behavioral reactions may be
attempts to ®nd alternative paths to goal achievement, withdrawal from eorts to achieve
organizational goals (turnover, absenteeism), interpersonal hostility or aggression, or organ-
izational aggression. Of particular concern is organizational aggression, which is any behavior
intended to hurt the organization, whether overt (strikes, work slowdowns, or lawsuits) or covert
(sabotage, withholding of output, or theft) (Spector, 1978).
The key to the frustration±aggression relationship is the mediation of aective response to the
antecedent situation. At the same time, however, a number of studies have established the impact
on this frustration±aective-behavioral response sequence of personality characteristics such as
aective traits and perceptions of control. For example, individual dierences in perceptions of
control of rewards have been studied in terms of both direct and moderating eects on the work
frustration±aggression sequence. Work Locus of Control refers to the extent to which people
believe they control reinforcements at work. Internality is a belief that individuals control their
own rewards; externality is a belief that others or luck control rewards (Spector and O'Connell,
1994). Locus of Control may impact both the experience of frustration and the behavioral
responses to experienced frustration. Spector and O'Connell found that internals experienced
lower levels of job stressors and work anxiety. Storms and Spector (1987) found some support for
the role of Work Locus of Control as a moderator in the frustration±behavioral reaction
relationship, particularly for sabotage. Individuals with low perceived control (externals) were
more likely to respond counterproductively to organizational frustration. Similarly, Perlow and
Latham (1993) found individuals with higher levels of externality were more likely to behave
abusively toward clients at work. These ®ndings may be explained by Allen and Greenberger's
(1980) suggestion that persons with lowered perceived control may attempt to modify their
environment and increase their feelings of control through destructive acts.
The role of aective dispositions has also been explored. Negative Aectivity (NA) is a
generalized dispositional tendency for an individual to experience negative emotions across time
and situations, even in the absence of aversive environmental events. The role of negative
aectivity has been studied with respect to perceptions of job stressors and constraints (Spector
and O'Connell, 1994) and self-reported frustrators and outcomes (Chen, 1989). NA is closely
related to Spielberger's construct of trait anxiety, which is de®ned as the relatively stable tendency
to perceive a wide variety of stressful situations as dangerous or threatening. Furthermore,
individuals high in trait anxiety are likely to respond to such perceived threats with higher and
more frequent elevations in state anxiety, consisting of both physiological arousal and perceived
feelings of tension, apprehension, nervousness, and worry (Spielberger and Sydeman, 1994).
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
918 S. FOX AND P. E. SPECTOR
Thus individuals high in trait anxiety would be expected both to perceive more organizational
stressors (situational constraints) and to respond to them with more frequent and intense
negative aective responses.
A second potentially important personality characteristic is trait anger, the likelihood that
individuals perceive a wide range of situations as anger-provoking. Individuals high in trait anger
are more likely to experience the emotional state of anger (subjective feelings of irritation,
annoyance, fury or rage, and associated physiological arousal) whenever they encounter frustrat-
ing or annoying conditions (Spielberger, Krasner and Solomon, 1988; Spielberger and Sydeman,
1994). Deenbacher (1992) found that individuals high in trait anger reported experiencing more
frequent and intense day-to-day anger across a wide variety of provocative situations, greater
anger in ongoing personally provocative situations, stronger tendencies to respond to provoca-
tions with physical and verbal antagonism, and lower instances of constructive coping. Thus
employees high in trait anger might be expected to respond to a wider range of organizational
constraints with behavioral expressions of anger.
In developing the State-Trait Anger Scale (STAS), Spielberger and colleagues identi®ed two
correlated factors comprising trait anger. Angry Temperament (T-Anger/T) describes individual
dierences in the general disposition to experience anger without specifying any provoking
circumstances. Individuals high in T-Anger/T are quick-tempered, express anger with little
provocation, and are often impulsive and more likely to express their anger outwardly than to
suppress it. Angry Reaction (T-Anger/R) describes angry reactions to speci®c situations involving
frustration or criticism. Individuals high in T-Anger/R are sensitive to perceived aronts and
negative evaluations, and respond with intense feelings of anger, although they are equally likely
to express or suppress their anger (Spielberger et al., 1995; Spielberger and Sydeman, 1994). Thus
we might expect individuals high in T-Anger/R to experience strong aective responses to
situational constraints, but not necessarily be more likely to express their anger behaviorally. On
the other hand, individuals high in T-Anger/T might be expected to respond to feelings of
frustration with counterproductive organizational behavior.
Current study
This study used structural equation modeling as well as zero-order correlational analysis to
evaluate a model of work frustration±aggression, as outlined in Figure 1. The hypothesized
relationship between organizational frustrated events (situational constraints) and counter-
productive behavioral responses is expected to be mediated by aective responses (frustration and
Figure 1. Model of work frustration±aggression
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 919
job dissatisfaction) to the experienced frustration. In addition, personality characteristics includ-
ing locus of control, trait anxiety and trait anger, are expected to substantially impact the
constraints±frustration±behavior sequence. In particular, individuals high in trait anxiety and
with an external work locus of control are expected to report higher levels of frustration and job
dissatisfaction. Individuals high in trait anger are likely to report more frequent incidences of
counterproductive work behavior, particularly individuals high in trait angry temperament
(T-Anger/T). The evaluation of the likelihood of being punished is also expected to impact the
aective-behavioral response relationship. Those individuals who perceive themselves as able to
harm the organization without getting caught and punished are more likely to report engaging in
counterproductive behaviors. In the current study, these relationships were tested using question-
naire data.
Hypotheses
Four hypotheses were tested in evaluating the model.
(1) Work-related frustrated events (situational constraints) will be associated with a range of
negative aective responses. A variety of job-related situational constraints in areas such
as training, resources, planning, assistance from others, time pressures, policies, red tape,
authority, and access to information will be positively associated with experienced
frustration and negatively associated with job satisfaction.
(2) Aective response to frustrated events (situational constraints) will be associated with
control beliefs and the personality predisposition to experience anxiety. Individuals with
external locus of control will report higher levels of frustration and job dissatisfaction, as
will individuals high in trait anxiety.
(3) Aective response to frustration will be related to counterproductive behavioral
responses. Lower levels of job satisfaction and higher levels of frustration will be
positively associated with a higher incidence of counterproductive work behaviors.
(4) Counterproductive behaviors will also be associated with individual dierences in prone-
ness to experience and express anger. Individuals with high levels of trait anger, in
particular angry temperament, will report a higher incidence of counterproductive beha-
vioral response. Similarly, the incidence of reported counterproductive behavior will be
higher for those individuals who believe they could ®nd a way to hurt the company
without being caught and punished.
Method
Participants
Participants were 185 full-time employees from eight corporations in Florida and Illinois. The
minimum quali®cation for inclusion in the study was that each subject used computers as a
required part of her/his task activity on the job. The participants represent a wide variety of
white-collar jobs, ranging from minimal computer expertise such as data entry to high technical
expertise, such as systems programmer-analysts. Of the respondents, 68.2 per cent were female
while 31.8 per cent were male; 51.9 per cent worked in private industry while 48.1 per cent were
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
920 S. FOX AND P. E. SPECTOR
Table 1. Sample size and response rate by source
Source Surveys distributed Surveys returned Response rate
Large national bank 40 25 62.5%
Telecommunications 40 22 55.0%
Large southeastern public university 200 89 44.5%
Exports 10 6 60.0%
Healthcare 17 7 41.2%
Utility 57 21 36.8%
Large midwestern private university 13 7 53.9%
Library 10 8 80.0%
Total 387 185 47.8%
Table 2. Summary of measures
Construct/scale (source) Number of items Range
Frustrated events/stressors
Job Eectiveness SurveyÐsituational constraints 13 1±5
(adaptation of O'Connor et al., 1982)
Belief of control
Work Locus of Control (Spector, 1988) 16 1±6
Disposition
State-Trait Personality Inventory (Spielberger, 1979)
Trait Anxiety scale 10 1±4
Trait Anger scale 10 1±4
Trait-Anger/Temperament 4 1±4
Trait-Anger/Reaction 4 1±4
Aective response
Frustration scale (Peters and O'Connor, 1980) 3 1±6
Job Satisfaction scale (Cammann et al., 1979) 3 1±6
Behavioral response
Job Reactions Survey (Spector, 1975; adaptations based 27 1±6
on Robinson and Bennett, 1995)
Likelihood of not being punished 2 1±6
state university employees. Participation in the study was on a strictly anonymous and voluntary
basis. Table 1 presents a breakdown of subjects and response rates by source.
Measures
Table 2 summarizes the characteristics and sources of measures included in the survey.
Situational constraints were measured by a shortened version of Eulberg, Peters, O'Connor
and Watson's (1984) Job Eectiveness Survey (JES), which was a modi®cation of O'Connor,
Peters, Rudolf and Pooyan's (1982) situational constraints scale. The JES was reduced to 13 items
to eliminate redundancy. The ®ve response choices range from `not at all accurate' to `completely
accurate'. High scores represent high levels of situational constraints on job eectiveness. Chen
and Spector (1992) reported the internal consistency to be 0.87 for the situational constraints
scale.
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 921
The 16-item Work Locus of Control Scale (Spector, 1988) addresses generalized beliefs about
control of rewards at work. The six response choices range from `disagree very much' to `agree
very much'. Half the items are written in an internal direction, representing a belief that indi-
viduals control their own rewards; the remaining items are written in an external direction,
representing a belief that others or luck control rewards (Spector and O'Connell, 1994). High
scores on the scale represent externality. Spector (1988) reported a Cronbach alpha of 0.80.
Aective disposition was measured with the Trait Anxiety (T-Anxiety) and Trait Anger
(T-Anger) scales of Spielberger's (1979) State-Trait Personality Inventory. The 10-item
T-Anxiety scale is used to measure a generalized dispositional tendency for an individual to
experience anxiety across time and situations. The 10-item T-Anger scale assesses the likelihood
that individuals perceive a wider range of situations as anger-provoking. For both the T-Anxiety
and T-Anger scales, the four response choices range from `almost never' to `almost always'. High
scores represent high levels of trait anxiety or trait anger, Spielberger (1979) reported alpha
coecients ranging from 0.80 to 0.92 for Trait Anxiety and 0.82 to 0.92 for Trait Anger, across
dierent ages and genders.
In addition, the T-Anger scale was broken down into two subscales of four items each,
measuring two related but distinct factors: Angry Temperament (T-Anger/T) and Angry
Reaction (T-Anger/R). T-Anger/T items, such as `I am a hotheaded person', describe gen-
eral disposition to express anger, independent of speci®c provocation. T-Anger/R items, such as
`It makes me furious when I am criticized', describe angry reactions to speci®c situational
provocations.
Job satisfaction was measured with a 3-item scale derived from the Michigan Organizational
Assessment Scale (Cammann, Fichman, Jenkins and Klesh, 1979). Spector and colleagues found
the internal consistency to range from 0.83 to 0.89 (Storms, 1985). Feelings of frustration were
measured by a 3-item scale from Peters and O'Connor (1980). Peters and O'Connor reported
internal consistency to be 0.81. For both the satisfaction and frustration scales, the six response
choices range from `disagree very much' to `agree very much'. After reverse-coding one item on
each scale, high scores represent high levels of satisfaction and frustration, respectively.
Counterproductive work behaviors were measured by a scale adapted from the Job Reactions
Survey (JRS) (Spector, 1975). For the purpose of this study, the JRS was adapted to more
fully sample the four domains delineated in Robinson and Bennett's (1995) typology of organ-
izational deviance. The four dimensions are production deviance (minor organizational),
property deviance (serious organizational), political deviance (minor interpersonal) and personal
aggression (serious interpersonal). The six response choices range from `never' to `extremely
often'. High scores represent high incidence of counterproductive behaviors.
The individual's assessment of the likelihood of being punished for engaging in counter-
productive behaviors was measured with two items. One item refers to the individual's ability to
generally hurt the company or mess up things at work without being caught; the second item
speci®es harming computer data or systems at work. The six response choices range from
`disagree very much' to `agree very much'. High scores represent a high level of con®dence in
being able to hurt the company without being caught.
Procedure
A survey booklet, letter of introduction and instructions, and return envelope were distributed to
each respondent between July 1995 and April 1996. The introductory letter emphasized that
participation was voluntary and anonymous. Respondents from the bank returned the completed
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
922 S. FOX AND P. E. SPECTOR
booklets to a sealed envelope on site. Respondents from the public university, export ®rm, and
library mailed their completed booklets directly to the researcher. Respondents from the
telecommunications, utility, and healthcare corporations and the private university were given
the choice of returning their booklets to a centrally located sealed box or envelope, or mailing the
survey directly back to the researcher. Of 387 distributed booklets, 185 usable surveys (47.8 per
cent) were returned.
Results
Descriptive statistics
Descriptive statistics on study variables in the model are presented in Table 3. Included are
sample sizes, means, standard deviations, possible ranges, observed ranges, and coecient
alphas.
Table 3. Descriptive statistics on study variables
Variable N M S.D. Possible range Actual range Coecient alpha
SIT 185 25.9 9.6 13±65 13±58 0.89
LOC 185 42.7 11.4 16±96 18±70 0.85
STPANX 183 16.8 4.4 10±40 10±36 0.81
STPANG 183 16.7 4.3 10±40 10±29 0.79
SAT 185 14.8 3.5 3±18 4±18 0.85
FRUST 184 9.3 3.7 3±18 3±18 0.76
BEH 185 36.9 8.2 25±150 25±77 0.86
PUN 185 4.4 3.5 1±12 1±12 0.88
Table 4 presents the numbers and percentages of survey respondents reporting having experi-
enced each of the 13 situational constraints at work. These numbers represent respondents who
did not characterize the statement as `Not at all accurate'.
Table 5 presents the numbers and percentages of survey respondents reporting having done
each of the 27 counterproductive behaviors at work. These numbers represent respondents who
did not answer `Never' for each item. Table 5 also shows the breakdowns of the behaviors into
Robinson and Bennett's (1995) dimensions of organizational deviance. Of interest is the low
base-rate occurrence of some of the possible behavioral responses. In particular, no respondent
reported having done two of the behaviors: (1) Purposely damaged a valuable piece of property
or equipment belonging to your employer. (2) Physically attacked a co-worker. These items
represent the extremes of organizational aggression, and their zero base rate in a sample of this
size and nature is not surprising. These two items were dropped in further analyses of the data.
Relationships among variables
Table 6 contains correlations among the variables in the model. As presented in Table 6,
situational constraints are highly correlated (p 5 0.001) to personality and control measures, as
well as aective and behavioral responses. Personality and control measures as well as aective
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 923
Table 4. Incidence of situational constraints reported
Item Description No. of Percentage
individuals of sample
SIT1 I often must depend on others who are not well trained 126 68%
SIT2 I have not been given enough training to do my job properly 100 54%
SIT3 I am not provided with the right materials and supplies to do my job 76 41%
SIT4 The tools and equipment I need to do my job are frequently 71 38%
not available
SIT5 Lack of time frequently keeps me from getting my job done 135 73%
SIT6 I am frequently given unscheduled activities to work on which keep 133 72%
me from getting my job done
SIT7 It is dicult to get others to provide the help required to do my job 116 63%
SIT8 Bad physical conditions (too cold, too bright, etc.) often interfere 58 31%
with getting my work done
SIT9 I frequently receive inconsistent policies, procedures and 113 61%
instructions, which makes it dicult to do my job
SIT10 Too much `red tape' frequently interferes with getting my 130 70%
work done
SIT11 I often have to follow the instructions of others even though I am 100 54%
in a better position to know what should be done
SIT12 I often do not have the information I must have to do my job when 110 60%
it is needed
SIT13 My department's budget does not allow me to get my job done 67 36%
responses are strongly related (p 5 0.001) to behavioral responses. Finally, likelihood of no
punishment is strongly (p 5 0.001) related to behavioral responses.
While most of the study variables show high correlations, a path analysis shows that the
pattern of relationships strongly supports the hypothesized model. This analysis was carried out
with LISREL 8 (Joreskog and Sorbom, 1993). Figure 2 shows standardized path coecients for
the model. Situational constraints are associated with aective response, in particular frustration.
Both high trait anxiety and external locus of control are associated with aective response.
Frustration is associated with counterproductive behaviors, thus supporting the mediating role
of aective response between situational constraints and behavioral responses. Finally, both
trait anger and likelihood of punishment show strong relationships to behavioral response.
Figure 2. Path analysis of work frustration±aggression model
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
924 S. FOX AND P. E. SPECTOR
Table 5. Incidence of counterproductive behaviors reported
Class of Item Description No. of Percentage
deviance individuals of sample
Minor organizational BEH1 Purposely wasted company materials/ 55 30%
supplies
BEH2 Daydreamed rather than did your work 165 89%
BEH3 Purposely ignored your boss 98 53%
BEH4 Complained about insigni®cant things 151 82%
at work
BEH5 Told people outside the job what a lousy 91 49%
place you work at
BEH6 Purposely did your work incorrectly 3 2%
BEH7 Felt good when something went wrong 39 21%
BEH8 Seriously considered quitting your job 136 74%
BEH9 Purposely came to work or came back from 101 55%
lunch breaks late
BEH10 Stayed home from work and said you were 95 51%
sick when you were not
BEH11 Purposely did not work hard when there were 82 44%
things to be done
Serious organizational BEH12 Purposely damaged a valuable piece of 0 0%
property or equipment belonging to
your employer
BEH13 Purposely littered or dirtied your place of 4 2%
work or your employer's property
BEH14 Taken any kind of drug at work to get high 7 4%
(including alcohol)
BEH15 Stolen something from work 38 21%
BEH16 Tried to cheat your employer 13 7%
Minor personal BEH17 Failed to help a co-worker 64 35%
BEH18 Withheld work-related information from 40 22%
a co-worker
BEH19 Played a practical joke on someone at work 116 63%
BEH20 Purposely interfered with someone else doing 22 12%
their job
BEH21 Started or continued a damaging or harmful 31 17%
rumor at work
BEH22 Blamed co-workers for errors that you made 19 10%
Serious personal BEH23 Started an argument with someone at work 59 32%
BEH24 Been nasty to a fellow worker 82 44%
BEH25 Stole something that belonged to a co-worker 1 1%
BEH26 Verbally abused a co-worker 26 14%
BEH27 Physically attacked a co-worker 0 0%
Overall, the model shows an excellent goodness of ®t (GFI 0.98; CFI 0.98; NFI 0.97;
RMSEA 0.078; chi-square (5 df) 10.41, p 4 0.05), with all paths signi®cant.
In addition, the counterproductive behavior items were broken down into Robinson and
Bennett's (1995) typology of organizational deviance. The behaviors are categorized as serious or
minor deviance, directed at organizational or personal targets (SERORG, MINORG, SERPER,
and MINPER). Table 7 presents the correlations of these behavioral response subscales with the
other major study variables. Finally, in order to increase our understanding of the role played by
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 925
Table 6. Correlations among organization-level study variables
SIT LOC ANX ANG TAR TAT SAT FRUST PUN BEH
SIT 1
LOC 0.31* 1
ANX 0.32* 0.35* 1
ANG 0.26* 0.31* 0.32* 1
TAR 0.29* 0.33* 0.33* 0.89* 1
TAT 0.16{ 0.13 0.16{ 0.72* 0.36* 1
SAT ÿ0.28* ÿ0.42* ÿ0.35* ÿ0.21{ ÿ0.19{ ÿ0.15{ 1
FRUST 0.54* 0.41* 0.41* 0.32* 0.37* 0.13 ÿ0.41* 1
PUN 0.15{ 0.07 0.15{ 0.17{ 0.13 0.17{ ÿ0.20{ 0.11 1
BEH 0.36* 0.30* 0.36* 0.59* 0.48* 0.43* ÿ0.37* 0.35* 0.39* 1
* p 5 0.001; { p 5 0.01; { p 5 0.05.
Table 7. Correlations between counterproductive behaviors and the situational, personality and aect
variables
SERORG MINORG SERPER MINPER ORG PER
SIT 0.05 0.39* 0.22{ 0.25* 0.37* 0.26*
LOC 0.17{ 0.32* 0.17{ 0.17{ 0.32* 0.19{
ANX 0.17{ 0.40* 0.24{ 0.14 0.40* 0.20{
ANG 0.27{ 0.57* 0.48* 0.44* 0.57* 0.50*
TAR 0.23{ 0.48* 0.37* 0.30* 0.48* 0.36*
TAT 0.19{ 0.38* 0.37* 0.40* 0.39* 0.42*
PUN 0.17{ 0.40* 0.30* 0.25* 0.39* 0.30*
SAT ÿ0.27* ÿ0.44* ÿ0.18{ ÿ0.09 ÿ0.45* ÿ0.14
FRUST 0.12 0.38* 0.20{ 0.22{ 0.36* 0.23{
SERORG 1
MINORG 0.43* 1
SERPER 0.31* 0.63* 1
MINPER 0.31* 0.59* 0.64* 1
ORG 0.56* 0.99* 0.63* 0.59* 1
PER 0.34* 0.67* 0.83* 0.94* 0.67* 1
Behaviors are broken down into the four categories of serious organizational, minor organizational, serious personal,
and minor personal (Robinson and Bennett, 1994). In addition, behaviors are broken down into two categories of
organizational and personal behaviors. Trait Anger is broken down into the two subscales of T-Anger/Reaction and T-
Anger/Temperament. * p 5 0.001; { p 5 0.01; { p 5 0.05.
anger in the model, Table 7 also presents the relationships among the deviance categories and the
two components of trait anger (T-Anger/Temperament and T-Anger/Reaction).
Of the counterproductive behaviors, minor organizational deviance shows the most consistent
pattern of strong relationship (p 5 0.001) with situational constraints, personality and control,
and aective responses. The situation±MINORG relationship is signi®cantly stronger than the
situation±SERORG, situation±SERPER, and situation±MINPER relationships (Hotelling
dependent t comparisons with William's correction 4.73, p 5 0.001; 2.94, p 5 0.01; and 2.22,
p 5 0.05, respectively). The satisfaction±MINORG relationship is signi®cantly stronger than the
satisfaction±SERORG, satisfaction±SERPER and satisfaction±MINPER relationships (Hotel-
ling dependent t comparisons with William's correction ÿ2.42, p 5 0.05; ÿ4.65, p 5 0.001; and
ÿ5.91, p 5 0.001, respectively). The frustration±MINORG relationship is signi®cantly stronger
than the frustration±SERORG, frustration±SERPER, and frustration±MINPER relationships
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
926 S. FOX AND P. E. SPECTOR
(Hotelling dependent t comparisons with William's correction 3.53, p 5 0.001; 2.88, p 5 0.01;
and 2.51, p 5 0.05, respectively). The serious organizational and the personal categories of
behavior are related to most of the situational, aect, and personality variables, but to a far lesser
extent.
Discussion
The major purpose of this study was to test a causal model of work frustration±aggression,
whereby organizational frustrated events (situational constraints) were related to counter-
productive organizational behaviors, mediated by aective responses to frustration. Disposi-
tional and belief variables (trait anxiety, trait anger, locus of control, and belief in the ability to
harm the organization without being punished) were included in the model. Using structural
equation modeling, the plausibility of this model was strongly supported, with goodness of ®t
indices of 0.97 and 0.98, and all paths signi®cant. In addition, speci®c correlational hypotheses
were tested.
Summary of tests of hypotheses
As predicted by hypothesis 1, work-related frustration events (situational constraints) were found
to be signi®cantly related to each of the aective variables in the model (Table 6). Situational
constraints were negatively correlated at the p 5 0.001 level of signi®cance with job satisfaction
(r ÿ0.28) and positively correlated at the p 5 0.001 level of signi®cance with frustration
(r 0.54).
As predicted by hypothesis 2, aective responses to frustrated events were associated with
control beliefs and anxiety proneness. As seen in Table 6, the highest correlates of satisfaction are
locus of control (r ÿ0.42, p 5 0.001) and trait anxiety (r ÿ0.35, p 5 0.001). Both locus of
control and trait anxiety correlate at r 0.41, p 5 0.001 with frustration.
As predicted by hypothesis 3, counterproductive behavioral responses were signi®cantly
related to aective responses. Highly signi®cant ( p 5 0.001) relationships were found between
behavioral responses and job satisfaction (r ÿ0.37) and between behavioral responses and
frustration (r 0.35). It is noteworthy that, in addition to the hypothesized mediation of the
situation±behavior relationship by aect, a signi®cant direct relationship was found between
situational constraints and behavioral responses (r 0.36, p 5 0.001).
As predicted by hypothesis 4, the incidence of counterproductive behaviors was also associated
with individual dierences in proneness to experience and express anger, as well as with the
employee's assessment of likelihood of no punishment for deviant behavior. The highest
correlates of behavioral response are trait anger (r 0.59, p 5 0.001) and likelihood of no
punishment (r 0.39, p 5 0.001). The results suggest that the combination of a personality prone
to anger and the belief that one has the ability to harm the organization without being punished
may be the strongest predictor of deviant behavior. It must be noted, however, that belief in no
punishment may be a result rather than a cause of having performed harmful behaviors (and not
having been caught)!
The possibility that antecedent variables contribute dierentially to serious versus minor
counterproductive behaviors targeting the organization versus people was tested. It is, however,
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 927
Figure 3. Exploratory path analysis with subcomponents of behavior and anger
possible that the lower correlations of situational constraints, frustration, and satisfaction with
serious organizational and serious personal behaviors might be attributable to the methodo-
logical artifact of lower base rates of the serious behaviors.
Finally, the two subscales of trait anger, angry temperament (T-Anger/T) and angry reaction
(T-Anger/R) were looked at in terms of dierential impact on counterproductive behavior.
Contrary to expectations, angry temperament did not correlate more highly with behavior than
did angry reactionÐboth components of trait anger were strongly associated with behavior
(r 0.43 and 0.48, respectively). However, it appears that angry temperament is more strongly
linked to behavioral responses targeting persons, while angry reaction is linked to behaviors
targeting the organization (see Table 7). This reinforces the call for further investigation into
dierential sources of interpersonal versus organizational aggression. In addition, fear of
punishment is associated with behaviors targeting the organization, but not personal aggression.
This is understandable in light of the wording of the items, involving one's ability to harm the
organization without being punished. A model distinguishing between organizational and
personal behaviors as well as between angry temperament and angry reaction was evaluated
with another path analysis (see Figure 3). This path analysis, while exploratory, supports the
plausibility of such a dierentiation (GFI 0.97; CFI 0.95; NFI 0.93; RMSEA 0.099;
chi-square (12 df) 33.09, p 5 0.01), with all paths signi®cant except for the path between
frustration and personal deviance. Although the path between frustration and organizational
deviance overshadows that between frustration and personal deviance, frustration and personal
behaviors nevertheless correlate signi®cantly. The stronger link to organizational deviance may
perhaps be attributed to the nature of the items measuring frustration as well as situational
constraints: the wording refers speci®cally to frustrations and constraints in the work organiza-
tion. Note that the path from T-Anger/R to personal and from T-Anger/T to organizational
deviance were nonsigni®cant and were trimmed from the model.
Self-report methodology
Several other factors pertaining to the present study should be considered when interpreting the
results. The most serious concern is the reliance in this study on self-report methodology. It is an
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
928 S. FOX AND P. E. SPECTOR
ongoing concern in organizational behavior research that the use of a single source of data,
particularly self-report questionnaires, may result in an overstatement of relationships among the
variables. However, there are three issues that might be addressed when considering self-report
methodology. (1) How appropriate is self-report for measuring the particular constructs of
interest in the context of the particular study? (2) Would alternative measures do a better job of
measuring those constructs, or come up with dierent results? (3) Are there practical and/or
ethical considerations that would prohibit the use of alternative measures and thus result in the
neglect of the study of important issues?
The model of frustration±aggression investigated in the current study is the association between
situational constraints experienced by the employee and counterproductive behavioral reactions
to the experienced frustration, mediated by emotional responses. Objective constraints and job
characteristics would seem to be of less importance to the cognitive, aective, and behavioral
processes, than the employees' perceptions and appraisals of those situational constraints and job
characteristics (Spector, 1997). As it is these perceptions, rather than objective reality that would
be expected to lead to experienced frustration and the associated aective and behavioral
responses, self-report measures may more appropriately tap critical features of the situation than
would more objective, non-incumbent measures. Where the research goal is understanding how
people view, feel about, and respond to their jobs, cross-sectional self-report methodology may be
most useful (Howard, 1994; Schmitt, 1994; Spector, 1994).
On the behavioral outcomes side, several researchers have pointed out the diculty of
obtaining hard criterion data for ¯agrant delinquent acts such as theft or property damage, while
individuals are still employed (Hogan and Hogan, 1989; Sackett and Harris, 1984). Furthermore,
hard criterion data (such as information in personnel ®les, supervisor reports, and turnover and
absenteeism records) suer both criterion de®ciency and contamination. They can represent only
those counterproductive behaviors in which the employee has been caught, which probably
represent a small subset of those behaviors of which the employee is aware and may report. At the
same time, a considerable body of literature has dealt with the diculties in pinpointing
situational antecedents of withdrawal behaviors such as turnover and absenteeism, which are
associated with multiple factors such as labor market conditions (Carston and Spector, 1987;
Kirschenbaum and Weisburg, 1994; Muchinsky and Morrow, 1980) and work/nonwork issues
(Youngblood, 1984). It is possible that counterproductive behaviors most relevant to a situa-
tional constraints±frustration±aggression model may be most appropriately accessed through
self-report measures.
Where either self-report or behavioral measures are appropriate, it may not necessarily be the
case that method variance associated with self-report will in¯ate the relationships of interest. Lee
(1993) summarizes research showing that people tend to underreport on questions about sensitive
topics such as deviant behavior, for fear of being caught and punished. This would result in an
attenuation of relationships, due to restriction of range and underreporting. On the other hand, it
is possible that equivalent results may be found using either self-report or hard behavioral
criteria, when such criteria are available. For example, Perlow and Latham (1993) looked at the
relationship between employee locus of control and aggressive counterproductive behavior by
workers caring for institutionalized mentally retarded clients. The behavioral criteria were
recorded instances of client abuse. They found a signi®cant correlation of 0.15 between locus of
control (external) and abuse. Storms and Spector (1987), using exclusively self-report method-
ology, found a signi®cant correlation of 0.16 between work locus of control (external) and
counterproductive behavioral reactions to organizational frustration.
Finally, there are practical and ethical reasons for relying on anonymous questionnaires in this
type of research. More important than the inherent diculty of gaining access to organizational
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 929
participants and con®dential personnel data, is an overriding ethical concern in doing organ-
izational research. Employees may rightly feel at risk in revealing behaviors which could result in
loss of job or even prosecution (Lee, 1993). Any methodology that could result in the identi®-
cation of respondents who have committed counterproductive behaviors, and thus endanger
their livelihoods or result in intensi®ed surveillance or punitive measures as a result of the
research, violates the most fundamental principles of doing ethical research in organizations.
Anonymous self-report seems most likely to avoid these ethical pitfalls.
Further considerations
A separate issue is the likelihood that the results of this study substantially underestimate both
the rate of counterproductive behaviors and the relationship between those behaviors and situa-
tional constraints. While the response rate (48 per cent of employees who received the question-
naire) was satisfactory, the small number of companies (8) which allowed distribution of the
questionnaire relative to the number of companies approached (19) may very well represent a self-
selecting factor. The primary reason given by companies in declining to participate was that they
were currently undergoing downsizing, restructuring, or other major changes, and that they did
not want to further `unsettle' their employees. Employees in precisely such companies might
provide the richest information about counterproductive behavioral responses to aversive work
environments.
This study merely scratches the surface of the issues related to counterproductive behavioral
responses to frustrating work environments. The construct anger might be more ®nely tuned by
measuring proclivity to express anger, rather than the more general trait anger (see the work of
Spielberger and colleagues on anger expression; Spielberger et al., 1995). A second area needing
clari®cation is the dierential patterns of relationships between the two measures of aective
response, frustration and satisfaction. The measurement of aective response to the work
environment might be more sensitively measured with a comprehensive aect-based scale (as
opposed to the more attitudinal measure of job satisfaction) such as the Job-Related Aective
Well-Being Scale (Van Katwyk, Spector, Fox and Kelloway, 1995).
Of particular interest would be to pursue the dierent patterns that surfaced for personal
versus organization forms of deviance. What types of external situational factors (home life,
work/nonwork role con¯ict) might account for personal as opposed to organizational aggres-
sion? Are there personality variables (such as nonspeci®c angry temperament) that might account
for an individual's choice of personal versus organizational behavioral response to frustration at
work? Finally, it would be interesting to break down personal aggression based on the targets of
the behaviors: supervisors, peers, subordinates, and clients. Do these classes of personal
aggression relate to dierent types of situational constraints?
On a broader level, further research is needed to tie in the organizational frustration±
aggression model with other currents of research investigating counterproductive behavioral role
choices, such as organizational justice, psychological contracts, availability of grievance pro-
cedures, and survivor syndrome in organizations ravaged by downsizing. Finally, it would be
useful to investigate the conditions under which behaviors an organizational deems counter-
productive, such as whistle-blowing and withdrawal, might in fact be adaptive from the point of
view of the well-being of the employees involved.
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
930 S. FOX AND P. E. SPECTOR
References
Allen, V. L. and Greenberger, D. B. (1980). `Destruction and perceived control'. In: Baum, A. and Singer,
J. E. (Eds) Applications of Personal Control, Erlbaum, Hillsdale, NJ.
Brief, A. P. and Motowidlo, S. J. (1986). `Prosocial organizational behaviors', Academy of Management
Review, 10, 710±725.
Cammann, C., Fichman, M., Jenkins, D. and Klesh, J. (1979). `The Michigan Organizational Assessment
Questionnaire'. Unpublished manuscript, University of Michigan, Ann Arbor.
Carston, J. M. and Spector, P. E. (1987). `Unemployment, job satisfaction, and employee turnover: A meta-
analytic test of the Muchinsky model', Journal of Applied Psychology, 72, 374±381.
Chen, P. Y. (1989). `Relationships of frustrators with aective, behavioral, and physical reactions:
The eects of potential moderators'. Masters Thesis, University of South Florida, Tampa, Florida.
Chen, P. Y. and Spector, P. E. (1992). `Relationships of work stressors with aggression, withdrawal, theft
and substance use: An exploratory study', Journal of Occupational and Organizational Psychology, 65,
117±184.
Deenbacher, J. L. (1992). `Trait anger: Theory, ®ndings, and implications'. In: Spielberger, C. D. and
Butcher, J. N. (Eds) Advances in Personality Assessment, Vol. 9, Lawrence Erlbaum Associates, Hillsdale,
NJ, pp. 177±201.
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H. and Sears, R. R. (1939). Frustration and Aggression,
Yale University Press, New Haven.
Eulberg, J. R., Peters, L. H., O'Connor, E. J. and Watson, T. W. (1984). Situational Performance
Constraints: A Selected Review of Relevant Literature (AFHRL-TP-83-48), Air Force Human Resources
Laboratories, Brooks AFB, San Antonio, TX.
Fishbein, M. and Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and
Research, Addison-Wesley, Reading, MA.
Hogan, J. and Hogan, R. (1989). `How to measure employee reliability', Journal of Applied Psychology, 74,
273±279.
Howard, G. S. (1994). `Why do people say nasty things about self-reports?' Journal of Organizational
Behavior, 15, 399±404.
Joreskog, K. G. and Sorbom, D. (1993). LISREL 8 User's Reference Guide, Scienti®c Software
International, Inc., Chicago.
Kirschenbaum, A. and Weisburg, J. (1994). `Predicting worker turnover: An assessment of intent on actual
separations', Human Relations, 43, 829±847.
Lee, R. M. (1993). Doing Research on Sensitive Topics, Sage, London.
McLean Parks, J. and Kidder, D. L. (1994). ` ``Till death us do part''. Changing work relationships in the
1990s'. In: Cooper, C. L. and Rousseau, D. M. (Eds) Trends in Organizational Behavior, John Wiley &
Sons, Ltd., Chichester, UK.
Muchinsky, P. M. and Morrow, P. C. (1980). `A multidisciplinary model of voluntary employee turnover',
Journal of Vocational Behavior, 17, 263±290.
O'Connor, E. J., Peters, L. H., Rudolf, C. J. and Pooyan, A. (1982). `Situational constraints and employee
aective reactions: A partial ®eld replication', Group and Organizational Studies, 7, 418±428.
Organ, D. W. (1990). `The motivational basis of organization citizenship behavior', Research in Organ-
izational Behavior, 12, 43±72.
Perlow, R. and Latham, L. L. (1993). `Relationship of client abuse with locus of control and gender:
A longitudinal study in mental retardation facilities', Journal of Applied Psychology, 78, 831±834.
Peters, L. H. and O'Connor, E. J. (1980). `Situational constraints and work outcomes: The in¯uences of a
frequently overlooked construct', Academy of Management Review, 5, 391±397.
Robinson, S. L. and Bennett, R. J. (1995). `A typology of deviant workplace behaviors: A multi-
dimensional scaling study', Academy of Management Journal, 38, 555±572.
Sackett, P. R. and Harris, M. M. (1984). `Honesty testing for personnel selection: A review and critique',
Personnel Psychology, 37, 221±245.
Schmitt, N. (1994). `Method bias: the importance of theory and measurement', Journal of Organizational
Behavior, 15, 393±398.
Spector, P. E. (1975). `Relationships of organizational frustration with reported behavioral reactions of
employees', Journal of Applied Psychology, 60, 635±637.
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
10991379, 1999, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/(SICI)1099-1379(199911)20:6<915::AID-JOB918>3.0.CO;2-6 by Egyptian National Sti. Network (Enstinet), Wiley Online Library on [16/03/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WORK FRUSTRATION±AGGRESSION 931
Spector, P. E. (1978). `Organizational frustration: A model and review of the literature', Personnel
Psychology, 31, 815±829.
Spector, P. E. (1988). `Development of the Work Locus of Control Scale', Journal of Occupational
Psychology, 61, 335±340.
Spector, P. E. (1994). `Using self-report questionnaires in OB research: A comment on the use of a
controversial method', Journal of Organizational Behavior, 15, 385±392.
Spector, P. E. (1997). `The role of frustration in antisocial behavior at work'. In: Jiacalone, R. A. and
Greenberg, J. (Eds) Anti-Social Behavior in Organizations, Sage, Thousand Oaks, CA, pp. 1±17.
Spector, P. E. and O'Connell, B. J. (1994). `The contribution of personality traits, negative aectivity, locus
of control and Type A to the subsequent reports of job stressors and job strains', Journal of Occupational
and Organizational Psychology, 67, 1±11.
Spielberger, C. D. (1979). `Preliminary manual for the State-Trait Personality Inventory (STPI)'.
Unpublished manuscript, University of South Florida, Tampa.
Spielberger, C. D. and Sydeman, S. J. (1994). `State-Trait anxiety inventory and State-Trait anger
expression inventory'. In: Maruish, M. E. (Ed.) The Use of Psychological Tests for Treatment Planning
and Outcome Assessment, LEA, Hillsdale, NJ.
Spielberger, C. D., Krasner, S. S. and Solomon, E. P. (1988). `The experience, expression, and control of
anger'. In: Janisse, M. P. (Ed.) Health Psychology: Individual Dierences and Stress, Springer Verlag/
Publishers, New York.
Spielberger, C. D., Reheiser, E. C. and Sydeman, S. J. (1995). `Measuring the experience, expression, and
control of anger'. In: Kassinove, H. (Ed.) Anger Disorders: De®nitions, Diagnosis, and Treatment, Taylor
& Francis, Washington, DC.
Storms, P. L. (1985). `Locus of control as a moderator of organizational frustration'. Masters Thesis,
University of South Florida, Tampa, Florida.
Storms, P. L. and Spector, P. E. (1987). `Relationships of organizational frustration with reported
behavioural reactions: The moderating eect of locus of control', Journal of Occupational Psychology, 60,
227±234.
Van Katwyk, P., Spector, P. E., Fox, S. and Kelloway, K. (1995). `How does my job make me feel?
The development of the Job-Related Aective Well-Being Scale (JAWS)'. Presentation at Society of
Industrial and Organizational Psychologists Conference.
Youngblood, S. A. (1984). `Work, nonwork, and withdrawal', Journal of Applied Psychology, 69, 106±117.
Copyright # 1999 John Wiley & Sons, Ltd. J. Organiz. Behav. 20, 915±931 (1999)
You might also like
- 100% (2)Immediate download (Ebook) Counterproductive work behavior: Investigations of actors and targets by Suzy Fox, Paul E. Spector, Lynne M. Andersson, Karl Aquino, Julian Barling, Robert A. Baron, Rebecca J. Bennett, Robert J. Bies, Susan M. Burroughs, Stale Einarsen, Robert Folger ISBN 9781591471653, 1591471656 ebooks 202481 pages
