INTERNATIONAL JOURNAL OF SELECTION AND ASSESSMENT VOLUME 10 NUMBERS 1/2 MARCH/JUNE 2002

Building an Integrative Model of Extra Role

Work Behaviors: A Comparison of
Counterproductive Work Behavior with
Organizational Citizenship Behavior
Donald E. Miles* Walter E. Borman
University of Florida, and Institute University of South Florida
of Human Performance, and Personnel Decisions Research
Decision-making, and Cybernetics Institute

Paul E. Spector Suzy Fox

University of South Florida Loyola University, Chicago

Researchers have recently drawn attention to performance elements that extend beyond
task performance (often labeled as organizational citizenship behavior, or OCB). A
second stream of research concerns detrimental, potentially destructive acts that hurt
colleagues or organizations (labeled counterproductive work behavior, or CWB).
Spector and Fox (in press) have proposed a model based upon conceptual parallels that
may help connect these two disparate fields of research, using positive and negative
emotions as intermediary variables between environmental conditions and extra-role
work behaviors. The current study tested portions of the overall model, and found
reasonable support for the components. Suggestions for further research designed to
test the overall model are presented.

Introduction aggression (Newman and Baron 1998; Spector 1975),

deviance (Holliger 1986; Robinson and Bennett 1995),

I n recent years there has been increasing interest in

exploring performance-related behaviors that go
beyond the assigned tasks and responsibilities for which
retaliation (Skarlicki and Folger 1997) and revenge (Bies,
Tripp and Kramer 1997). Recently, Spector and Fox (in
press) developed a model that integrates both OCB and
employees are typically held accountable. On the positive CWB, showing both parallels and distinctions in the
side, researchers have looked at voluntary behavior that environmental and personal factors hypothesized to be
goes beyond the core tasks under the rubrics of precursors to both.
organizational citizenship behavior or OCB (Organ The Spector and Fox (in press) model describes how
1988; Smith, Organ and Near 1983) and contextual environmental and personal factors lead to behavior
performance (Borman and Motowidlo 1993). On the through mediating processes of perception and emotion.
negative side, there is an independent stream of research The process described by the model begins with the work
concerning detrimental, potentially destructive acts that environment. As people go about their work lives, events
hurt colleagues or organizations, which are labeled and situations at work provide stimuli that are perceived
counterproductive work behavior or CWB (Fox, Spector and appraised. Some of these events and situations are
and Miles 1999; Spector and Fox, in press), workplace perceived/appraised in such a way that they induce
emotional reactions, either positive or negative. Both
* Address for correspondence: Donald E. Miles, Department of
forms of emotions produce action tendencies and
Psychology, University of South Florida, 4202 E. Fowler Ave., BEH intentions to act. Negative emotions tend to produce
339, Tampa, Fl. 33620. e-mail: [email protected] CWB and positive emotions tend to produce OCB. There

ß Blackwell Publishers Ltd 2002, 108 Cowley Road, Oxford OX4 1JF, UK and
350 Main Street, Malden, MA 02148, USA. 51
52 DONALD E. MILES, WALTER E. BORMAN, PAUL E. SPECTOR AND SUZY FOX

are two additional factors that are important in this Empirical evidence exists supporting hypothesized links
process. First, perceptions of control affect the extent to with personality. Trait anger and trait anxiety (Spielberger
which emotions are experienced in response to workplace 1979) have been shown to be related to CWBs (Fox and
events. Situations seen as controllable are more likely to Spector 1999; Fox et al. 1999). Furthermore, Ones,
induce positive emotion and situations seen as Viswesvaran and Schmidt (1993) found that integrity tests
uncontrollable are more likely to induce negative emotion. that apparently assessed three of the `Big Five' personality
Furthermore, control perceptions in combination with dimensions (Costa and McCrae 1988), conscientiousness,
emotions enhance the likelihood that OCB (high control) agreeableness, and emotional stability, were related to
or CWB (low control) will occur. Second, personality CWB. Empathy has been shown to predict helping behavior
characteristics play an important role in both OCB and in a number of settings (e.g. Allen 1999; Davis 1996;
CWB, although for the most part it seems that different Eisenberg and Miller 1987; Penner et al. 1995). Organ and
characteristics are involved for each. Affective dispositions Ryan (1995) found that conscientiousness, agreeableness,
such as state anger and anxiety, neuroticism, and positive affectivity, and negative affectivity have been
emotional experience are related to emotional response related to OCB. Borman, Penner, Allen and Motowidlo
and are therefore related to CWB. Locus of control is also (2001) reported that other personality variables, including
related to CWB. On the other hand, OCB is related to locus of control, collectivism, and personal initiative are
empathy and helpfulness, a construct devised by Penner, related to OCB.
Fritzsche, Craiger and Freifeld, (1995) that consists of a Support for the idea that CWB and OCB can be
willingness to help while maintaining emotional distance oppositely affected by the same factors comes from Hunt
and feeling high ability to help. (1996) who investigated relations among nine dimensions
The model suggests that the causal connections among of both positive and negative performance-related
these variables is neither simple nor unidirectional. behaviors for 18,146 employees from 36 companies.
Background emotional state affects perceptions, so that The positive dimensions all related negatively with the
an individual already angry or anxious will be more negative dimensions, with correlations ranging from
likely to perceive a situation in such a way that it will ÿ.11 to ÿ.73. Additionally, Robinson and O'Leary-Kelly
induce even more emotion. CWB and OCB feed back (1998) found that their measure of CWB correlated ÿ.24
into emotions ± performing a negative act may induce and ÿ.46 with two different measures of OCB.
anxiety and guilt where performing something positive
may induce positive emotions of pride and accom-
plishment. Whereas control perceptions affect per- Current Research Project
ceptions of situations, these do not occur in a vacuum
and the situation itself provides cues concerning control. This research was designed to investigate important
Finally, behaviors themselves affect the environment, and linkages in the model proposed by Spector and Fox (in
employees can find themselves caught in loops where a press), and to explore within the same study both CWB and
situation induces an emotion that results in a situation OCB. The general aim is to investigate how environmental
that induces even more of that emotion. For example, a conditions, trait anger, and positive and negative affect
person who engages in OCB might receive appreciation relate to OCB and CWB respectively. If these key linkages
and recognition that induces positive emotion and a are supported, then further research would be warranted to
greater likelihood of repeating the OCB. test the overall model more completely.
Spector and Fox (in press) discuss specific job
conditions that are likely to induce emotion and
subsequent behavior. Organizational constraints are Method
situations at work that inhibit or limit task performance
(Peters and O'Connor 1980). Results from various meta-
Participants
analyses (Spector 1997; Spector and Jex 1998) have been
interpreted as showing linkages between various Participants were 203 workers who were employed in
constraints and feelings of frustration and anxiety, as various manufacturing, financial, utility, entertainment,
well as to CWB (Chen and Spector 1992; Fox et al. 1999). and academic organizations, and were also enrolled at the
Interpersonal conflict (getting into arguments with University of South Florida. Of the participants, 65 (32%)
coworkers and poor treatment by coworkers and were men and 138 (68%) were women. Some 63 respon-
supervisors) and being the victim of abusive behavior dents (31%) worked 40 or more hours per week in their job.
also have been shown to be related to negative emotions The remaining 140 respondents (68%) worked between 20
(Spector and Jex 1998) and to CWB (Chen and Spector and 39 hours per week. Average age for the workers was 25
1992; Frone 1998). Workload has been shown to relate to years of age, with a range of 17 to 53. Some 66% of the
negative emotion (Spector and Jex 1998) and to a small workers were white, 18% were of African-American origin,
extent to CWB (Chen and Spector 1992). and 12% were of Hispanic or Latin American origin.

International Journal of Selection and Assessment ß Blackwell Publishers Ltd 2002

BUILDING AN INTEGRATICE MODEL OF EXTRA ROLE WORK BEHAVIORS 53

Measures arguments, yelling, and rudeness in interactions with co-

workers. The five response choices for these items range
The anonymous self-report survey included measures
from 1 = `never' to 5 = `always'. For this scale, high
designed to assess a variety of psychological constructs
scores indicate high levels of interpersonal conflict at
as well as work behaviors. Psychological constructs that
work.
were assessed included conditions in the work environ-
ment (workload, constraints, and conflict), personality
Organizational Citizenship Behavior (OCB). Component
characteristics (trait anger), and affect (positive emotions,
items from Coleman and Borman's (2000) Citizenship
negative emotions). Work behaviors assessed included
Performance Scale were used to assess OCB. The original
organizational citizenship behaviors (OCB) and counter-
twelve-item scale is based upon a conceptual model
productive work behaviors (CWB).
postulating that organizational citizenship is composed
Personality characteristics. The State-Trait Personality
of three components: personal support (support for
Inventory (Spielberger 1979) was used to measure
individuals), organizational support (support for the
affective disposition of anger. The 10-item T-Anger scale
organization), and conscientious initiative (persisting
was designed to assess the likelihood of perceiving a wide
with extra effort despite difficult conditions). However,
range of situations as anger-provoking. Four response
previous research has not supported the three factors
choices range from 1 = `almost never' to 4 = `almost
(Sun 2001). Additionally, this scale was originally
always', with high scores representing high levels of trait
designed to be filled out by supervisors who assessed
anxiety or trait anger.
the performance of direct reports. For this study, both
the directions (self-report) and the scaling have been
Affect. The Job-Related Affective Well-Being Scale
changed in order to be consistent with the directions and
(JAWS), developed by Van Katwyk, Fox, Spector, and
the scaling used for assessing CWB. The five response
Kelloway (2000), was used to assess a wide range of
choices range from 1 = `never' to 5 = `every day'.
emotions experienced in response to conditions of the
Finally, for this study, only six of the original items were
job. Employees respond to items on the JAWS by
used (the other six items were essentially CWBs that
indicating how often any part of the job has made them
would be reverse scored). A sample item is `Cooperated
feel each of 30 emotional states. The five response
fully with others by willingly sacrificing your own
choices range from 1 = `almost never' to 5 = `extremely
personal interests for the good of the team'.
often or always'. High scores represent high levels of
An additional nine-item scale had been developed
each emotion. A negative emotions score was obtained
specifically for this study in order to assess the relative
by summing scores on the 17 negative affect items; a
incidence of OCB. Sample items include `Suggested ideas
positive emotions score was obtained by summing the
for improvement' and `Completed assigned work ahead
scores on the 13 positive affect items.
of schedule'. The five response choices range from 1 =
`never' to 5 = `every day'. The items from the two scales
Environmental characteristics. The Organizational
were combined to create an OCB scale with 15 items.
Constraints Scale or OCS (Spector and Jex 1998) was
However, due to the results of a preliminary scaling
used to measure work constraints. This 11-item scale was
study, one item from the Borman and Coleman scale was
based on constraint areas identified by Peters and
dropped, and one of the newly developed items was also
O'Connor (1980). Respondents indicated the frequency
dropped, due to low item-to-total correlations. The
with which their job performance was hindered by
reliability (Cronbach's Generalized Alpha) of the revised
constraints such as availability of resources, co-workers,
13-item scale was .87 based on the results from the
interruptions, rules and procedures, and inadequate
preliminary study (N = 110).
training. The five response choices range from 1 = `Less
than once per month or never' to 5 = `Several times per
Counterproductive Work Behavior (CWB). Counter-
day'. High scores represent high levels of constraints.
productive work behaviors were assessed with a
The five-item Quantitative Workload Inventory
behavioral checklist based on a master list compiled from
(Spector and Jex 1998) was used to assess the quantity
a number of existing measures (Fox and Spector 1999;
of work, the amount of free time, and the extent to which
Hollinger 1986; Knorz and Zapf 1996; Neuman and Baron
an employee was required to work hard and fast. The five
1998; Robinson and Bennett 1995; Skarlicki and Folger
response choices for these items range from 1 = `never' to
1997; Spector 1975). The authors' goal was to avoid
5 = `always'. High scores for this scale represent high
duplication while including as many distinct behaviors as
workload.
possible. We put the resulting 45-item list into a scale
In order to assess the perception of interpersonal
format, asking participants to indicate how often they had
conflict by the workers, the Interpersonal Conflict at
done each of the behaviors on their present job. The five
Work Scale (Spector and Jex 1998) was used. This four-
response choices range from 1 = `never' to 5 = `every
item scale was designed to measure the frequency of
day', with high scores representing high incidence of

ß Blackwell Publishers Ltd 2002 Volume 10 Numbers 1/2 March/June 2002

54 DONALD E. MILES, WALTER E. BORMAN, PAUL E. SPECTOR AND SUZY FOX

Table 1. Means, standard deviations, reliabilities and correlation coefficients of study variables

Variable M SD 1 2 3 4 5 6 7 8

1. Constraints 1.98 .71 (.86)

2. Conflict 1.49 .70 .41 (.72)
3. Workload 3.25 1.14 .47 .34 (.88)
4. Positive emotion 2.95 .78 -.23 -.30 -.27 (.93)
5. Negative emotion 2.19 .71 .54 .52 .54 -.50 (.92)
6. Trait anger 1.76 .50 .28 .25 .21 -.22 .38 (.83)
7. CWB 1.42 .29 .32 .19 .21 -.22 .35 .38 (.87)
8. OCB 3.46 .70 .26 .24 .31 .23 .12 .03 ÿ.11 (.90)

Notes: N = 203; coefficient alphas are shown in parentheses on the diagonal

Correlations above |.14| are significant at the p < .05 level
Correlations above |.18| are significant at the p < .01 level

counterproductive behaviors. Sample items include: inversely with positive emotions (r = -.22). Negative
`Insulted someone about their job performance'; `Made emotions and positive emotions are inversely correlated,
fun of someone's personal life'; `Refused to help a with r = -.50. OCB and CWB are inversely correlated,
coworker'; `Started an argument with a coworker'. with r = -.11 (which is non significant at p < .05).
Two separate regression analyses were conducted with
OCB and CWB as dependent measures, respectively.
Results Each regression featured five predictor variables, which
included three environmental conditions, as well as trait
Preliminary analyses included evaluating the reported anger and affect. For the first regression, with OCB as the
frequency of occurrence of several behaviors (primarily criterion, the three environmental conditions (workload,
CWB). Several behaviors were reported as never constraints, and interpersonal conflicts) were entered as a
occurring by at least 90% of the respondents and were block of variables, followed by trait anger, then followed
dropped from the later analyses. This left a total of 33 by positive emotions. Results of this regression procedure
counterproductive work behaviors that were used in the are shown in Table 2. The three environmental
subsequent analyses. As stated previously, items from the conditions explained 11% of the variance of OCB, with
two OCB scales were combined, to create a thirteen-item the beta weights of both workload and interpersonal
scale dealing with organizational citizenship behaviors. conflict significant at p < .05. Trait anger did not add
Table 1 provides intercorrelations, reliabilities, means any unique variance. Then positive emotion added an
and standard deviations for the primary variables. additional 13% of explained variance.
Environmental conditions correlate with negative For the second regression, with CWB as the criterion,
emotions (ranging from r = .52 to r = .54). Environ- the three environmental conditions (workload, con-
mental conditions also correlate (inversely) with positive straints, and interpersonal conflicts) were entered as a
emotions (ranging from r = -.23 to r = -.30). Trait anger block of variables, followed by trait anger, then followed
correlates with negative emotions (r = .38), and by negative emotions. The three environmental conditions

Table 2. Regression analysis predicting Organizational Citizenship Behaviors

Variable t Adj. R2
R2

Workload .19 3.97 **

Constraints .14 1.83
Conflict .23 3.03 ** .11

Trait anger ÿ.05 ÿ.46 .11 .00

Positive emotion .39 5.99 ** .24 .13

Notes: Values after all variables entered

* p < .05; ** p < .01

International Journal of Selection and Assessment ß Blackwell Publishers Ltd 2002

BUILDING AN INTEGRATICE MODEL OF EXTRA ROLE WORK BEHAVIORS 55

Table 3. Regression analysis predicting Counterproductive Work Behaviors

Variable t Adj. R2
R2

Workload ÿ.01 ÿ.18

Constraints .06 2.00 *
Conflict ÿ.01 ÿ.47 .09

Trait anger .16 4.09 ** .18 .09

Negative emotion .07 2.04 * .19 .01

Notes: Values after all variables entered

* p < .05; ** p < .01

explained 9% of the variance of CWB, with the beta This phenomenon may also be occurring with
weight for constraints significant, with t (197) = 2.00, p < perceptions of interpersonal conflict in the workplace.
.05. The three environmental conditions explained 9% of There is a positive relationship with negative emotions,
the variance of CWB, with the beta weights of work as well as with CWB. Perceptions of interpersonal
constraints significant, with p < .05. Trait anger then conflict are inversely related to positive emotions, but
added an additional 9% of explained variance, with then are positively related to OCB. Workers may view
negative emotion adding an additional 1% of variance. conditions under which there are higher levels of conflict
The beta weight for trait anger was significant with t (197) as creating (from their perspective) the opportunity to
= 4.09, p < .01. The beta weight for negative emotions overcome negative conditions, persevere, and go above
was also significant with t (197) = 2.04, p < .05. The and beyond the call of duty.
results from this regression analysis are shown in Table 3. Support for the basic premise that environmental
conditions (as perceived by the worker) lead to emotional
reactions, which can then influence behavior has
practical value to employers. An approach that manages
Discussion emotion can be helpful in encouraging positive voluntary
behaviors. This same approach can also be helpful in
There does seem to be reasonable support for some of the
inhibiting negative forms of behavior.
component parts of the overall model, as proposed by
Spector and Fox (in press). This model proposed that
environmental conditions (as perceived by the worker)
Limitations of the Study
relate to emotional reactions, which can then influence
behavior. Perceptions of the work environment relate to This study was conducted using cross-sectional, self-
positive emotion, which is positively correlated with the report measures of workers in various jobs. This strategy
occurrence of OCB. Conversely, negative perceptions of can be viewed as problematical for a multitude of
the work environment relate to negative emotion, which reasons. This research, however, was specifically
is positively correlated with the occurrence of CWB. designed to focus on the perceptions of workers
Personality characteristics (specifically, trait anger) (especially regarding their perceptions of the work
provide additional explanatory value in interpreting environment) and on their affective responses to those
these relationships (especially regarding CWB). perceptions. Additionally, gathering uncontaminated
In reviewing the pattern of zero order correlations as data regarding CWB is also problematic (primarily
well as the two regression analyses for OCB and CWB, dealing with ethical issues). Therefore, for this study
complex relationships between environmental conditions, self-report methodology seemed both prudent and
personality characteristics, emotion, and behaviors appear reasonable. However, creative research designs (possibly
to exist. For example, perceptions of workload may lead including the identification of known groups: people,
to negative emotions, but not directly to CWB. However, jobs, or companies that historically feature high levels of
workload is inversely related to positive emotions, but CWB or OCB) still need to be developed and used.
then is positively related to OCB. High workload can be Privacy concerns and guarantees of anonymity will
stressful and perceived in a negative manner, but may also always be of paramount importance. Additionally,
create (at least in the worker's perception) the opportunity inclusion of some type of measure to detect individuals
to persevere, and continue to work in spite of difficult responding in a socially desirable manner would also be
conditions. appropriate.

ß Blackwell Publishers Ltd 2002 Volume 10 Numbers 1/2 March/June 2002

56 DONALD E. MILES, WALTER E. BORMAN, PAUL E. SPECTOR AND SUZY FOX

A second major limitation deals with the scope of the Organizational Psychology, Atlanta, 30 April±2 May.
study. Although this study was not designed to be a full- Bies, R.J., Tripp, T.M. and Kramer, R.M. (1997) At the
breaking point: Cognitive and social dynamics of revenge in
scale test of the proposed model, that objective remains organizations. In R.A. Giacalone and J. Greenberg (eds)
as a primary goal of our research. The inclusion of Antisocial Behavior in Organizations. Thousand Oaks, CA:
additional personality variables such as conscien- Sage.
tiousness, agreeableness, emotional stability, locus of Borman, W.C. and Motowidlo, S.J. (1993) Expanding the
control, collectivism, personal initiative, and empathy criterion domain to include elements of contextual
performance. In N. Schmitt and W.C. Borman (eds)
should provide additional information, but will require Personnel Selection. San Francisco: Jossey-Bass.
much larger samples, and careful choice or development Borman, W.C., Penner, L.A., Allen, T.D. and Motowidlo, S.J.
of psychometrically sound instruments. The use of (2001) Personality predictors of citizenship performance.
measures such as the Hogan Personality Inventory International Journal of Selection and Assessment, 9, 52±59.
(Hogan and Hogan 1995) and the Prosocial Personality Campbell, J.P. (1990) An overview of the army selection and
classification project (Project A). Personnel Psychology, 43,
Battery (Penner et al. 1995) would seem to be prudent. 231±239.
Additional environmental conditions that should be Campbell, J.P., Gasser, M.B. and Oswald, F.L. (1996) The
incorporated in future tests of the overall model include substantive nature of performance variability. In K.R.
procedural and distributive justice, perceptions of Murphy (ed.) Individual Differences and Behavior in
autonomy, and especially perceptions of control. Organizations. San Francisco: Jossey-Bass.
Campbell, J.P., McCloy, R.A., Oppler, S.H. and Sager, C.E.
A final limitation deals with the OCB and CWB (1993) A theory of performance. In N. Schmitt and W.C.
measures used in this study. The Coleman and Borman Borman (eds) Personnel Selection in Organizations. San
(2000) scale did not perform as expected, but when Francisco: Jossey-Bass.
selected items were combined with the additional eight Chen, P.Y. and Spector, P.E. (1992) Relationships of work
OCB items, did provide valuable information. Future stressors with aggression, withdrawal, theft and substance
use: An exploratory study. Journal of Occupational and
research, however, will need to be conducted using an Organizational Psychology, 65, 177±184.
improved scale (or multiple scales) with sound Coleman, V.I. and Borman, W.C. (2000) Investigating the
psychometric properties, that is consistent with broader underlying structure of the citizenship performance domain.
conceptualizations of OCB. Human Resource Management Review, 10, 25±44.
Additionally, the CWB behavior checklist needs to be Costa, P.T., Jr. and McCrae, R.R. (1988) From catalog to
classification: Murray's needs and the five-factor model.
continually updated and refined. For example, additional Journal of Personality and Social Psychology, 55, 258±265.
behaviors related to computers and computing need to be Davis, M.H. (1996) Empathy: A Social Psychological Approach.
added. Scale items addressing excessive use of e-mail and Boulder, CO: Westview.
Internet surfing are being pilot tested, and will be added Eisenberg, N. and Miller, P.A. (1987) Empathy and prosocial
to later versions of the checklist. behavior. Psychological Bulletin, 101, 91±119.
Fox, S. and Spector, P.E. (1999) A model of work frustration-
The general support for component parts of the model aggression. Journal of Organizational Behavior, 20, 915±
should provide support for the premise that similar 931.
processes may impact CWB and OCB. This ideally will Fox, S., Spector, P.E. and Miles, D. (1999) Counterproductive
lead to further research with larger samples and behavior (CPB) in response to job stressors and
additional variables to test the full model. The inclusion organizational justice: The moderator effect of autonomy
and emotion traits. Paper presented at the annual conference
of additional personality variables, including measures of the Society for Industrial and Organizational Psychology,
from `The Big Five' (Costa and McCrae 1988) should Atlanta, 30 April±2 May.
provide valuable information and help clarify some of the Hogan, J. and Hogan, R. (1995) Hogan Personality Inventory
more complex relationships within the model. manual, 2nd edn. Tulsa, OK: Hogan Assessment Systems.
Additionally, the inclusion of bi-directional relationships Hollinger, R.C. (1986) Acts against the workplace: Social
bonding and employee deviance. Deviant Behavior, 7, 53±75.
and mediated variables, will more appropriately repre- Hunt, S.T. (1996) Generic work behavior: An investigation into
sent the real world of work. Then, the Emotion ± the dimensions of entry-level, hourly job performance.
Voluntary Behavior Model may become a valuable Personnel Psychology, 49, 51±83.
addition within the multidimensional model of job Knorz, C. and Zapf, D. (1996) Mobbing ± eine extreme Form
performance proposed by Campbell and colleagues sozialer Stressoren am Arbeitsplatz [Mobbing ± an extreme
form of social stressors in the workplace]. Zeitschrift fuÈr
(Campbell 1990; Campbell, Gasser and Oswald 1996; Arbeits- und Organisationspsychologie, 40, 12±21.
Campbell, McCloy, Oppler and Sager 1993). Neuman, J.H. and Baron, R.A. (1998) Workplace violence and
workplace aggression: Evidence concerning specific forms,
potential causes, and preferred targets. Journal of
References Management, 24, 391-419.
Ones, D.S., Viswesvaran, C. and Schmidt, F.L. (1993)
Comprehensive meta-analysis of integrity test validities:
Allen, T.D. (1999) Mentoring others: Mentor dispositions and Findings and implications for personnel selection and
desired proteÂge characteristics. Paper presented at the theories of job performance [monograph]. Journal of
annual meeting of the Society for Industrial and Applied Psychology, 78, 679±703.

International Journal of Selection and Assessment ß Blackwell Publishers Ltd 2002

BUILDING AN INTEGRATICE MODEL OF EXTRA ROLE WORK BEHAVIORS 57

Organ, D.W. (1988) Organizational Citizenship Behavior: The with reported behavioral reactions of employees. Journal of
Good Soldier Syndrome. Lexington, MA: Lexington Books. Applied Psychology, 60, 635±637.
Organ, D.W. and Ryan, K. (1995) A meta-analytic review of Spector, P.E. (1997) The role of frustration in antisocial
attitudinal and dispositional predictors of organizational behavior at work. In R.A. Giacalone and J. Greenberg
citizenship behavior. Personnel Psychology, 48, 775±802. (eds) Antisocial Behavior in Organizations. Thousand Oaks,
Penner, L.A., Fritzsche, B.A., Craiger, J.P. and Freifeld, T.R. CA: Sage.
(1995) Measuring the prosocial personality. In J. Butcher Spector, P.E. and Fox, S. (in press) An emotion-centered model
and C.D. Spielberger (eds) Advances in Personality of voluntary work behavior: Some parallels between
Assessment, vol. 10. Hillsdale, NJ: Lawrence Erlbaum Counterproductive Work Behaviors (CWB) and
Associates. Organizational Citizenship Behavior (OCB). Human
Peters, L.H. and O'Connor, E.J. (1980) Situational constraints Resources Management Review.
and work outcomes: The influences of a frequently Spector, P.E. and Jex, S.M. (1998) Development of four self-
overlooked construct. Academy of Management Review, 5, report measures of job stressors and strain: Interpersonal
391±397. Conflict at Work Scale, Organizational Constraints Scale,
Robinson, S.L. and Bennett, R.J. (1995) A typology of deviant Quantitative Workload Inventory, and Physical Symptoms
workplace behaviors: A multidimensional scaling study. Inventory. Journal of Occupational Health Psychology, 3,
Academy of Management Journal, 38, 555±572. 356±367.
Robinson, S.L. and O'Leary-Kelly, A.M. (1998) Monkey see, Spielberger, C.D. (1979) Preliminary manual for the State-Trait
monkey do: The influence of work groups on the antisocial Personality Inventory (STPI). Unpublished manuscript,
behavior of employees. Academy of Management Journal, University of South Florida, Tampa.
41, 658±672. Sun, S. (2001) Predicting job satisfaction and organizational
Skarlicki, D.P. and Folger, R. (1997) Retaliation in the citizenship behavior with individualism-collectivism in PR
workplace: The roles of distributive, procedural, and China and the United States. Unpublished doctoral
interactional justice. Journal of Applied Psychology, 82, dissertation, University of South Florida, Tampa.
434±443. Van Katwyk, P.T., Fox, S., Spector, P.E. and Kelloway, E.K.
Smith, C.A., Organ, D.W. and Near, J.P. (1983) Organizational (1999) Using the Job-related Affective Well-being Scale
citizenship behavior: Its nature and antecedents. Journal of (JAWS) to investigate affective responses to work stressors.
Applied Psychology, 68, 653±663. Journal of Occupational Health Psychology, 5, 219±230.
Spector, P.E. (1975) Relationships of organizational frustration

ß Blackwell Publishers Ltd 2002 Volume 10 Numbers 1/2 March/June 2002