INTERNATIONAL JOURNAL OF SELECTION AND ASSESSMENT VOLUME 11 NUMBER 1 MARCH 2003

Investigating the Dimensionality of

Counterproductive Work Behavior
Melissa L. Gruys* and Paul R. Sackett
Washington State University University of Minnesota

The study investigated the dimensionality of counterproductive work behavior (CWB)

by examining the relationships between various counterproductive behaviors. Utilizing
a university alumni sample (N = 343), data was collected through both self-report and
direct judgments of the likelihood of co-occurrence. Eleven categories of CWB were
examined: (1) Theft and Related Behavior; (2) Destruction of Property; (3) Misuse of
Information; (4) Misuse of Time and Resources; (5) Unsafe Behavior; (6) Poor
Attendance; (7) Poor Quality Work; (8) Alcohol Use; (9) Drug Use; (10) Inappropriate
Verbal Actions; and (11) Inappropriate Physical Actions. CWB items and categories
were generally positively related. Multidimensional scaling analysis suggests that the
CWB categories vary on two dimensions: an Interpersonal-Organizational dimension
and a Task Relevance dimension.

Introduction CWB, the focus is on the behavior itself rather than on

the results or consequences of the behavior (e.g., the

O ne of the central themes within the field of

Industrial/Organizational Psychology is mapping
the relationship between individual characteristics (e.g.,
skills, abilities, personality traits) and workplace
behavior. Understanding these relationships permits
organizations to select individuals with characteristics
linked to desired workplace behaviors. Historically,
personnel selection research has focused primarily on
identifying candidates likely to engage in desirable work
behavior (e.g., those likely to produce work of high
quality, or to produce work at a high rate). More recently,
increased attention has been paid to an expanded array of
job behaviors, including the broad domain of undesirable,
counterproductive behaviors (Campbell 1990).
Robinson and Greenberg (1998) reviewed a wide
variety of definitions of counterproductive behavior that
are found in the literature. More recently, Sackett and
DeVore (2001) provide another thorough review. Based
on their treatment, we use the following definition of
counterproductive work behavior (CWB) in this
research: any intentional behavior on the part of an
organization member viewed by the organization as
contrary to its legitimate interests. With this definition of
* Address for correspondence: Melissa L. Gruys, Department of
Management and Decision Sciences, Washington State University,
14204 NE Salmon Creek Ave., Vancouver WA 98686. E-mail:
gruys@vancouver.wsu.edu
harm which is done). Only intentional behaviors are
included in this definition. While accidental actions may
cause harm, they are not the focus of the study. The
definition encompasses behavior that is targeted at both
individuals and at the organization, as both types of
actions can have severe consequences on the
organization. Behavior of organizational members is
included but the behavior of outsiders (e.g., clients.
former employees) is not included. While outsider actions
may be a concern, the actions of current employees are of
foremost concern to any organization.
The domain of counterproductive work behaviors is
quite broad. Prior to the early 1980s, there was a
considerable amount of research on individual
counterproductive behaviors in the workplace, such as
employee theft, pilferage, sabotage, slow and sloppy
performance, tardiness and absenteeism (e.g., Altheide,
Adler, Adler, and Altheide, 1978; Bensman and Gerver,
1963; Cressey, 1953; Gouldner, 1954; Henry, 1978a,
1978b; Horning, 1970; Mars, 1973; Robin, 1969, 1970;
Roy, 1952, 1959; Taylor and Walton, 1971). However,
due to the lack of an accepted framework or theory for
looking at such behaviors, these studies were each seen
as an investigation into a specific type of behavior rather
than as an effort to explore CWB in a more general
sense.
The key question this article addresses is `what is the
underlying structure of CWB?' The article looks at the

Blackwell Publishing Ltd 2003, 9600 Garsington Road, Oxford OX4 2DQ, UK and
30 350 Main Street, Malden, MA 02148, USA.
DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR
covariance of counterproductive behaviors in order to
examine the structure of such behavior.
Two important streams of relevant research are
outlined here. The first stream involves attempts to use
psychological tests to predict employee theft. Reviews by
Sackett and colleagues (Sackett and Decker, 1979; Sackett
and Harris, 1984; Sackett, Burris, and Callahan, 1989;
Sackett and Wanek, 1996) detail developments in this
area. The critical point is that integrity tests initially
developed to predict theft have over time been found to
predict a variety of counterproductive behaviors other
than theft, including absenteeism, violation of safety
rules, and drug and alcohol use on the job. However, this
work has been hindered by the lack of a clear picture of
the underlying structure of CWB.
The second stream involves attempts at the develop-
ment of taxonomies of counterproductive work beha-
viors. A seminal body of work examining a broad range
of counterproductive behaviors is that of Hollinger and
Clark (1982, 1983a, 1983b) who developed a broad list of
counterproductive behaviors, provided a conceptual
framework for interrelating those behaviors, and
collected self-report data from large employee samples
in three industries. They proposed that
counterproductive behaviors could be grouped into two
broad categories. The first category is `property
deviance', involving misuse of employer assets. Examples
include theft, property damage, and misuse of discount
privileges. The second category is `production deviance',
involving violating norms about how work is to be
accomplished. This includes not being on the job as
scheduled (absence, tardiness, long breaks) and behaviors
that detract from production when on the job (drug and
alcohol use, intentional slow or sloppy work). Empirical
work by Hollinger and Clark (1982, 1983a) has
supported the distinction between property and
production deviance.
Robinson and Bennett (1995) noted that the set of
behaviors examined by Hollinger and Clark did not
include interpersonal counterproductive behaviors, such
as sexual harassment, and set out to expand upon the
Hollinger and Clark framework. They had workers
generate a large number of critical incidents of
counterproductive behavior, obtained ratings of the
similarity between pairs of behaviors, and subjected the
resulting pairwise similarity matrix to multidimensional
scaling. They obtained a two-dimensional solution, with
one dimension differentiating behavior toward the
organization (Hollinger and Clark's production and
property deviance) from interpersonal behavior toward
other organizational members (e.g., harassment, gossip,
verbal abuse), and the other dimension representing a
continuum from minor to serious offenses. Arraying
behaviors in this two-dimensional space, Robinson and
Bennett labeled the resulting four quadrants as property
deviance (organizational serious), production deviance
Blackwell Publishing Ltd 2003
31
(organizational minor), personal aggression (inter-
personal serious, including behaviors such as harassment
and theft from co-workers), and political deviance
(interpersonal minor, including behaviors such as
favoritism, gossip, and blaming others for one's mistakes).
More recently, Bennett and Robinson (2000) have
developed a measure of workplace deviance and
presented construct validity evidence. This study
supported the Robinson and Bennett (1995) distinction
between production and property deviance. However,
the authors recommend dropping the serious to minor
dimension, as this dimension represents more of a
quantitative distinction than a qualitative one. A recent
review by Sackett and DeVore (2001) argues for a
hierarchical model of counterproductive behavior in
which there is a single general factor under which there
are two more specific factors represented by property and
production deviance. Within the two more specific
factors would fall even more specific types of behavior
(e.g., absenteeism, theft, etc.). Thus, research in the area
supports that counterproductive workplace behavior
likely has a general factor with more specific factors
representing organizational deviance and interpersonal
deviance beneath that.
While the current literature provides some insight into
the structure of CWB, there remain questions regarding
the dimensionality of such behavior. For example, within
either organizational deviance or interpersonal deviance,
are there various behaviors that are exhibited together
more frequently? How do the behaviors co-vary? One
approach to examining the dimensionality of counter-
productive behavior is multidimensional scaling, as
conducted by Robinson and Bennett (1995). A different
sample and a different approach to ratings would be a
good test of the structure of counterproductive behavior.
It is critical to note that the categorization scheme put
forth by Robinson and Bennett (1995) is based on
workers' perceptions of the similarity of pairs of
behaviors. When asked to judge the similarity of pairs
of behaviors. respondents were not constrained as to the
basis for their similarity judgments. The purpose of the
multidimensional scaling technique is to identify
dimensions underlying similarity judgments, and their
study findings indicate that, in the aggregate, respondents
used the two dimensions of organization as target vs.
other person as target and minor offense vs. serious
offense as the basis for their similarity judgments. While
it is useful to know what dimensions underlie perceptions
of similarity between counterproductive behaviors, we
posit here that a key issue for understanding inter-
relationships among various forms of counterproductive
behavior is the covariance of occurrence among these
behaviors. The question is whether individuals who
engage in one form of counterproductive behavior are
also likely to engage in others. It is possible that the
pattern of co-occurrence among counterproductive
Volume 11 Number 1 March 2003
32
behaviors is quite different from the pattern of perceived
similarity emerging from the Robinson and Bennett
perceptual similarity task. For example, theft from a co-
worker and verbal abuse of a customer are located close
together in the multidimensional space emerging from
Robinson and Bennett's analysis, in that both are serious
offenses that target another person. However, the two
behaviors may differ on a variety of other dimensions
(e.g., a public vs. private dimension, a planned vs.
unplanned dimension), making the rate of co-occurrence
of the two an open question.
The co-occurrence of counterproductive work
behaviors is of considerable interest in that an under-
standing of the interrelationships among these behaviors
can help guide subsequent work on antecedents of
counterproductive behaviors and on interventions aimed
at reducing the occurrence of counterproductive
behaviors. A finding of high rates of co-occurrence
among particular sets of counterproductive behaviors
may suggest common antecedents for these behaviors
and the possibility that an intervention aimed at one may
also influence others in the set. If, for example, all
counterproductive behaviors co-varied highly, resulting
in a single general factor, a common set of antecedents
for all of these behaviors might be implicated. At the
other extreme, a finding of low covariance might suggest
the need for separate research programs for each
counterproductive behavior. It should be noted, however,
that high intercorrelations between behaviors do not
necessarily mean that variance associated with specific
behaviors is unimportant.
How might one get insight into the rate of co-
occurrence of counterproductive behaviors on the part of
employees? The ideal would be the objective measure-
ment of each form of counterproductive behavior. We
acknowledge what has been viewed as the Achilles heel
of counterproductive behavior research, namely, that
while some forms of counterproductive behavior are
public (e.g., absence), many are acts by employees who
do not wish to be detected (e.g., theft, sabotage, harass-
ment). In the face of the difficulties of direct observation,
data on the covariance of counterproductive behaviors
might come from various sources, such as self-reports of
the rate of occurrence by employees themselves,
judgments by others (e.g., supervisors) of the rate of
occurrence, and direct judgments about the rate of co-
occurrence of counterproductive behaviors by employees.
As none of these strategies is clearly ideal, we make use
of dual strategies in the current research. First, we obtain
self-reports of the likelihood of engaging in various
counterproductive behaviors. Examining correlations
among the various behaviors will shed light on the
dimensionality issue. Second, we obtain direct judgments
about the likelihood of co-occurrence of pairs of
behavior categories, and transform these judgments into
a matrix of similarities between different categories of
International Journal of Selection and Assessment
MELISSA L. GRUYS AND PAUL R. SACKETT
counterproductive behavior. Thus, the study involved
two types of data: self-ratings and co-occurrence ratings,
and can be viewed as two parallel attempts to examine
the covariance among counterproductive behaviors. We
present these two types of data as Study 1 and Study 2
respectively.
Study 1: Self-Ratings of Counterproductive
Behavior
Sample
Alumni of a small Midwestern undergraduate liberal arts
institution were surveyed. This sample allowed for
inclusion of individuals from a wide variety of
occupations and industries who were employed in a
wide variety of organizations across the United States.
Surveys were mailed to 1,102 alumni and were returned
by 363 for a response rate of 33%. Ten of the surveys
arrived after the deadline and could not be included in
the analyses and ten additional surveys were eliminated
due to incomplete data. The resulting sample size for the
study was 343.
Some 94% of the sample were White, 56% were
female, 73% were married, and the mean age of
participants was 40 years (SD 10.15). The sample
was one in which the education level of participants was
extremely high, with 59% of the sample having education
beyond an undergraduate degree. Some 94% of the
sample were employed and 73% were employed full-
time.1 Some 86% of the participants were employed in
managerial, technical, and professional positions. The
sample had mean tenure in current job and job
experience of 8.37 years (SD 8.04) and 15.5 years
(SD 9.82), respectively, and a mean income of $55,400
(SD 57.39).
The survey described below was pilot-tested on 115
undergraduate students enrolled in management classes
at a large Midwestern university. The mean age of
participants was 21.7 years (with a range of 19 to 43);
55% were female, and 87% were White. Some 79% of
participants were employed. For those employed, mean
work hours per week was 21.5 (SD 9.46), and mean
tenure in present job was 18 months (SD 18.37).
Survey Construction
A list of over 250 counterproductive work behaviors was
compiled from the literature for use in the study. Keyword
searches were conducted using major databases (e.g.,
PSYCInfo, Academic Universe, etc.) to identify relevant
literature that may contain relevant behaviors. Literature
in the areas of psychology, management, business, and
sociology was examined. Items were obtained from 15
different articles.2 Redundant items were removed, and
Blackwell Publishing Ltd 2003
DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR
ultimately 66 behaviors were included in the study. The
behavior list was then sorted into 11 separate categories of
CWB based on the similarity of content. The following 11
categories of behaviors were formed: (1) Theft and
Related Behavior; (2) Destruction of Property; (3) Misuse
of Information; (4) Misuse of Time and Resources; (5)
Unsafe Behavior; (6) Poor Attendance; (7) Poor Quality
Work; (8) Alcohol Use; (9) Drug Use; (10) Inappropriate
Verbal Actions; and (11) Inappropriate Physical Actions.
Table 1 presents the CWB categories and the behaviors in
each category. The initial sort of the behaviors into the
categories was conducted by the first author. In a group
session eight doctoral students familiar with the research
project and its objectives independently sorted the items
into the categories. There were virtually no discrepancies
between judges, and thus no changes were made to the
initial item sort.
Since the study aimed to examine the structure of
CWB, the survey included a wide range of items. The
Bennett and Robinson (2000) measure of workplace
deviance was not used because the measure did not
represent a broad enough range of behaviors. Bennett
and Robinson acknowledge that in an effort to develop a
reliable and valid scale, they have included only the most
common forms of employee deviance. The 11 categories
of behavior used in the current study include the whole
range of behavior addressed by Bennett and Robinson
(2000). For many of the Bennett and Robinson items, on
the current survey, there were several more detailed items
related to that particular behavior.
Participants were asked to rate whether they would
engage in each of the counterproductive behaviors on a
seven-point scale with 1 being anchored with `No Matter
What the Circumstances, I Would Not Engage in the
Behavior' and 7 being anchored with `In a Wide Variety
of Circumstances, I Would Engage in the Behavior.'
Individuals were asked to respond according to the
assumption that they would have the opportunity in their
workplace situation to engage in all of the listed
behaviors. This type of rating avoids a situation in
which an employee may not have the opportunity to
engage in one of the behaviors being assessed. Thus, all
items are equally applicable to participants regardless of
the job in which they are employed. In addition, the
wording of the anchors on the rating scale attempted to
take into account the circumstances under which an
individual would engage in various behaviors. The
instructions for the survey prompted participants to
consider various circumstances when rating the items
(e.g., working conditions, penalties or punishments for
wrongdoing, unfair treatment, level of job satisfaction).
Thus the study did not ask for reports of behavior, but
rather of behavioral intentions. There is extensive
research documenting the very strong relationship
between intentions and behavior, much of it linked to
the Theory of Reasoned Action (Fishbein and Ajzen,
Blackwell Publishing Ltd 2003
33
1975; Fishbein, 1980), which posits intentions as the most
direct precursor to behavior.
In addition, the survey also included a section that
asked participants to provide demographic and
background information. The variables collected were:
age, sex, race, education, employment status (employed
or not), occupation and industry, tenure, and amount of
full-time work experience.
Analysis
The central analysis for Study 1 involved combining the
66 items into the 11 conceptually derived composites,
computing the intercorrelations among the 11 com-
posites, and conducting a principal components analysis
of this correlation matrix. Exploratory analysis was
done. as there was no strong a priori theory as to the
dimensionality of these 11 behavior dimensions.
Prior to these analyses, the internal consistency
reliability of the 11 proposed composites was examined,
and confirmatory factor analysis was conducted to
examine the proposed 11 dimensions. Absent strong
conceptual alternative hypotheses, we tested a single
factor model and the hypothesized 11- factor model.
It is acknowledged here that there is a potential for
confusion in Study 1 regarding the use of factor-analytic
procedures for two very different purposes. First, factor-
analytic procedures were used at the item level to test the
viability of combining the 66 items into the 11 categories
of behavior. Confirmatory factor analysis was used to
create and justify the use of the behavior categories.
Second, exploratory factor-analytic procedures were used
at the category level to carry out the main purpose of the
study, to examine the dimensionality of the 11 behavior
categories. This results in an unusual data analytic
pattern, namely, that we use confirmatory factor analysis
in the first stage of our analysis and exploratory factor
analysis in the second. This is intentional: the first stage
of the project involved determining whether a set of items
conformed to a hypothesized 11-factor solution (hence a
confirmatory strategy) while the second stage involved
examining the structure of the 11 categories (for which
we did not have strong a priori hypotheses and thus used
an exploratory strategy).
Results
Table 1 presents means and standard deviations for the
items for each CWB category. Mean responses ranged
from 1 to 4.7 on the seven-point scale and most items had
means of less than 3. These means indicate that
participants reported that they are not very likely to
engage in most of the behaviors included on the survey, a
result not surprising given the nature of the behaviors
Volume 11 Number 1 March 2003
34 MELISSA L. GRUYS AND PAUL R. SACKETT

Table 1: Descriptive statistics for counterproductive work behavior categories and items

Category and Items Number Alpha Composite Mean Std. Dev.

of items Estimate Mean

Theft and Related Behavior 10 .77 1.71

Help another person or advise them how to take 1.15 0.52
company property or merchandise.
Take cash or property belonging to the company. 1.66 1.12
Misuse business expense account. 1.55 0.99
Take cash or property belonging to a co-worker. 1.06 0.30
Take office supplies from the company. 2.76 1.65
Take petty cash from the company. 1.15 0.55
Take cash or property belonging to a customer. 1.06 0.25
Give away goods or services for free. 2.97 1.82
Provide goods or services at less than the price 1.89 1.39
established by the company.
Misuse employee discount privileges. 1.82 1.23

Destruction of Property 4 .66 1.11

Deface, damage, or destroy property, belonging 1.10 0.39
to a co-worker.
Deface, damage, or destroy property, belonging 1.07 0.29
to a customer.
Deface, damage, or destroy property, equipment, 1.17 0.47
or product belonging to the company.
Deliberately sabotage the production of product 1.09 0.44
in the company.

Misuse of Information 5 .71 1.57

Destroy or falsify company records or documents. 1.34 0.68
Discuss confidential matters with unauthorized 2.06 1.29
personnel within or outside the organization.
Intentionally fail to give a supervisor or co-worker 1.57 0.93
necessary information.
Provide the organization with false information to 1.20 0.51
obtain a job (i.e., regarding education or
experience).
Lie to employer or supervisor to cover up a mistake. 2.72 1.49

Misuse of Time and Resources 13 .90 2.81

Conduct personal business during work time. 3.41 1.80
Spend time on the internet for reasons not 3.47 1.99
related to work.
Take a long lunch or coffee break without approval. 3.12 1.98
Waste time on the job. 2.89 1.49
Waste company resources. 1.93 1.07
Use company resources you aren't authorized to use. 2.08 1.26
Make personal long distance calls at work. 2.86 1.85
Mail personal packages at work. 2.21 1.70
Make personal photocopies at work. 4.28 1.95
Use email for personal purposes. 4.66 2.17
Play computer games during work time. 2.14 1.58
Alter time card to get paid for more hours than 1.41 0.88
you worked.
Work unnecessary overtime. 2.02 1.57

Unsafe Behavior 4 .71 1.97

Endanger yourself by not following safety 1.94 1.15
procedures.

International Journal of Selection and Assessment Blackwell Publishing Ltd 2003

DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR 35

Endanger coworkers by not following safety 1.34 0.65

procedures.
Endanger customers by not following safety 1.31 0.62
procedures.
Fail to read the manual outlining safety procedures. 3.28 1.66

Poor Attendance 5 .77 2.06

Be absent from work without a legitimate excuse. 1.87 1.30
Intentionally come to work late. 1.86 1.32
Use sick leave when not really sick. 2.70 1.80
Leave work early without permission. 2.59 1.77
Miss work without calling in. 1.25 0.69

Poor Quality Work 3 .86 1.37

Intentionally perform your job below 1.34 0.76
acceptable standards.
Intentionally do work badly or incorrectly. 1.24 0.63
Intentionally do slow or sloppy work. 1.51 0.88

Alcohol Use 3 .59 1.35

Come to work under the influence of alcohol. 1.11 0.49
Have your performance affected due to a hangover 1.48 0.91
from alcohol.
Engage in alcohol consumption on the job. 1.44 1.06

Drug Use 4 .71 1.04

Engage in drug use on the job. 1.03 0.16
Come to work under the influence of drugs. 1.04 0.22
Possess or sell drugs on company property. 1.02 0.15
Have your performance affected due to a hangover 1.07 0.33
from drugs.

Inappropriate Verbal Actions 8 .82 1.83

Argue or fight with a co-worker. 2.89 1.41
Yell or shout on the job. 1.91 1.06
Verbally abuse a customer. 1.29 0.65
Verbally abuse a co-worker. 1.41 0.84
Verbally abuse a supervisor. 1.27 0.72
Use sexually explicit language in the workplace. 1.84 1.28
Argue or fight with a supervisor. 2.28 1.38
Argue or fight with a customer. 1.65 0.96

Inappropriate Physical Actions 7 .82 1.08

Physically attack (e.g., pushing, shoving, hitting) 1.11 0.33
a co-worker.
Physically attack (e.g., pushing, shoving, hitting) 1.04 0.20
a customer.
Physically attack (e.g., pushing, shoving, hitting) 1.06 0.27
a supervisor.
Make unwanted sexual advances toward a 1.11 0.49
subordinate.
Make unwanted sexual advances toward a 1.07 0.36
supervisor.
Make unwanted sexual advances toward a 1.10 0.44
co-worker.
Make unwanted sexual advances toward a 1.07 0.28
customer.

Note. The composite means are reported as a mean item for comparability.

Blackwell Publishing Ltd 2003 Volume 11 Number 1 March 2003

36
being rated. However, there were some behaviors which
sizable portions of the sample endorsed as being
behaviors in which they would be quite likely to engage,
such as `Use email for personal purposes' (item mean
4.66 on a seven-point scale), `Make personal photocopies
at work' (mean 4.28), `Use internet for non-work
purposes' (mean 3.47), and `Use sick leave when not
really sick' (mean 2.70). This result is consistent with
the findings of Bennett and Robinson (2000) who found
that for many of the counterproductive behavior items
surveyed, there was a relatively high participation rate.
Thus, it is not the case that all behaviors that at least
some organizations view as counterproductive are
uncommon.
We note that means and standard deviations for all
individual CWB items were very consistent between the
student pilot sample and the main study sample. The
correlation between the individual item means for the
two samples was .95 and the correlation between
individual item standard deviations was .92. These
findings are striking given how different the two samples
are in terms of age as well as work and life experience.
This suggests that the findings are not specific to the
alumni sample used here.
Table 1 presents the alpha estimates of reliability for
each of the CWB category composites. The alpha
estimates of reliability for the categories exceed .7, with
the exception of Destruction of Property and Alcohol Use
which have alpha estimates of .66 and .59, respectively.
As an additional check on the appropriateness of the
11 categories used here, confirmatory factor analyses
were conducted using LISREL. The first model that was
run in LISREL was a single factor model that specified
that all 66 of the items would load on a single
counterproductive behavior factor. This model produced
an extremely poor fit. The Goodness of Fit (GFI) index
was .50, the Adjusted Goodness of Fit (AGFI) was .46,
the Standardized Root Mean Square Residual (SRMR)
.093, and the Root Mean Square Error of Approximation
(RMSEA) was .12. An 11-factor model was run using
item parcels of 35 items per parcel (with items being
randomly assigned within category to a parcel). The fit of
an 11-factor model was substantially better with
goodness of fit statistics as follows: the GFI was .83,
the AGFI was .76, the SRMR was .083, and the RMSEA
was .064. All of the maximum likelihood estimates for
the paths were significant (p<.01). Though the fit of this
model was better, the fit was moderate.
Items making up each of the 11 dimensions were then
summed to form composites. Table 2 presents means,
standard deviations, and intercorrelations for the 11
dimensions and for other demographic variables.
Reliabilities are indicated on the diagonal. Given that
the number of items per CWB category varies, the means
for the category composites that are presented are the
mean per item. The highest means are for Misuse of
International Journal of Selection and Assessment
MELISSA L. GRUYS AND PAUL R. SACKETT
Time and Resources; Unsafe Behavior, and Poor
Attendance. The lowest means for the CWB categories
are those for Destruction of Property, Drug Use, and
Inappropriate Physical Actions.
Most of the background and demographic variables
were relatively uncorrelated with the 11 categories of
CWB, with the exception of age and work experience.
There were significant negative relationships between age
and six of the 11 categories. This indicates a trend of
older individuals being less likely to engage in these types
of counterproductive behavior. Work experience has a
similar negative relationship with the counterproductive
behavior categories. This is not surprising given the
correlation of .86 between age and work experience. In
general, these results are not unexpected given the results
of past research which suggests that older individuals and
those with longer tenure are less likely to engage in
counterproductive behavior in the workplace (e.g..
Hollinger and Clark, 1983b).
The correlations between the CWB categories ranged
from .17 to .71 and are all positive and significant at the
p<.01 level. The average correlation is .43. This is
similar to the correlation of .46 (uncorrected) between
organizational and interpersonal deviance found by
Bennett and Robinson (2000).
Principal components analysis was conducted on the
categories of behaviors using an oblique rotation. The first
two eigenvalues were 5.51 and 1.17. Thus a scree plot
suggests a 1-factor solution, while an eigenvalue-greater-
than-one rule of thumb suggests two factors. We examined
both solutions. The 2-factor solution produces two highly
correlated factors (r :50), and the second factor is
clearly methodological rather than substantive: factor 1
represents dimensions with high means and factor 2
represents dimensions with low means. Thus we focused
on the 1-factor solution, and present the results in Table 3.
All of the categories load quite highly on the first factor.
Thus, the results suggest a strong common dimension
underlying CWB. All categories load highly on the first
factor, which may be characterized as general
counterproductive work behavior. These results suggest
that as the likelihood of an individual engaging in a
certain type of CWB increases, the likelihood of that
individual engaging in a wide variety of other types of
CWB also increases.
Study 2: Co-occurrence Ratings of
Counterproductive Work Behavior
Sample
The same sample of 343 college alumni used in Study 1
was used in Study 2. Both the self-rating task described
above and the co-occurrence ratings described here were
included in the same mail survey.
Blackwell Publishing Ltd 2003
 Blackwell Publishing Ltd 2003

DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR

Table 2: Descriptive statistics and intercorrelations for all variables

Variable Mean1 SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

1. Theft 1.71 0.64 (.77)

2. Property 1.11 0.29 .47** (.66)
3. Informat 1.57 0.62 .71** .55** (.71)
4. Timereso 2.81 1.13 .74** .43** .71** (.90)
5. Safety 1.97 0.81 .46** .35** .57** .48** (.71)
6. Attend 2.06 1.04 .63** .38** .64** .76** .37** (.77)
7. Workqual 1.37 0.68 .55** .51** .64** .50** .33** .53** (.86)
8. Alcohol 1.35 0.63 .38** .30** .35** .40** .33** .29** .20** (.59)
9. Drugs 1.04 0.17 .27** .44** .29** .21** .17** .27** .38** .28** (.71)
10. Verbal 1.83 0.72 .55** .52** .57** .54** .47** .47** .42** .39** .27** (.82)
11. Physical 1.08 0.24 .42** .59** .42** .36** .24** .31** .26** .29** .38** .43** (.82)
12. Gender 1.44 0.50 .07 .18** .01 .00 .07 .06 .02 .12** .06 .12* .15**
13. Age 39.77 10.15 .15** .05 .18** .30** .08 .24** .12* .04 .09 .08 .11* .24**
14. Married 0.73 0.45 .04 .03 .07 .13* .07 .04 .11* .00 .02 .04 .09 .12* .27**
15. Educ. 17.74 2.05 .02 .06 .14* .07 .07 .06 .03 .01 .03 .09 .09 .12* .12* .11
16. Employed 0.93 0.25 .10 .06 .06 .13* .08 .08 .04 .02 .05 .07 .08 .12* .02 .04 .09
17. Orgtype 0.47 0.50 .08 .08 .07 .15** .00 .07 .06 .07 .05 .04 .08 .11* .00 .02 .56** .11
18. Tenure 8.36 8.09 .08 .00 .09 .15** .10 .10 .06 .04 .04 .03 .03 .26** .57** .22** .07 .11 .02
19. Hrs/Wk. 42.01 12.06 .06 .03 .10 .04 .08 .10 .06 .03 .04 .03 .04 .34** .03 .07 .17** .31** .13* .20**
20. Work 15.50 9.82 .13* .02 .19** .27** .05 .24** .14** .02 .07 .04 .07 .35** .86** .24** .11 .01 .04 .56** .21**
Exp
Volume 11 Number 1 March 2003

21. Income 55.38 57.39 .01 .03 .09 .01 .03 .01 .13* .02 .06 .01 .05 .26** .31** .17** .23** .11 .24** .22** .30** .35**

Notes:
1
For comparability, the means for category composites reflect the mean per item. Alpha estimates of reliability are along the diagonal when available.
Sample sizes range from N 318 to N 343 due to missing data. For variables 1 through 11, the counterproductive behavior categories, Theft Theft and Related Behavior; Property Destruction of Property;
Informat Misuse of Information; Timereso Misuse of Time and Resources; Safety Unsafe Behavior; Attend Poor Attendance; Workqual Poor Quality Work; Alcohol Alcohol Use; Drugs Drug Use; Verbal
Inappropriate Verbal Actions; Physical Inappropriate Physical Actions. Gender was a dichotomous variable which was coded as follows: 0 female and 1 male. Married was a dichotomous variable which was
coded as follows: 1 married and 0 not married. Education represents the total years of education. Employed was a dichotomous variable which was coded as follows: 0 not employed and 1 employed. Orgtype
represents the type of organization employed in and was coded as follows: 0 non-profit and 1 for profit. Tenure and work experience were measured in years. Hrs/Wk. represents the number of hours worked per
week. Income was measured in dollars. * p < :05. ** p < :01.

37
38
Table 3: Alumni sample factor loadings for principal
components analysis of counterproductive work
behavior categories
One Factor Solution
Category Factor Loading
Misuse of Information .87
Theft and Related Behavior .83
Misuse of Time and Resources .82
Poor Attendance .76
Inappropriate Verbal Actions .74
Destruction of Property .71
Poor Quality Work .70
Unsafe Behavior .63
Inappropriate Physical Actions .60
Alcohol Use .53
Drug Use .48
Note: N 343.
Survey Construction
Based on preliminary empirical support for the 11
categories of CWB from Study 1, the co-occurrence
rating task was performed using these categories. The
survey asked participants to rate the co-occurrence of
categories of CWB. This task involved the comparison of
pairs of categories. The survey described co-occurrence
of behaviors as follows:
If the behaviors in the two categories co-occur,
employees who engage in behaviors in one of the
categories would also engage in the behaviors in the
other category. Further, employees who would not
engage in behaviors in one of the categories would
also not engage in behaviors in the other category in
the pair. If the behaviors in the two categories do not
co-occur, then employees who engage in behaviors in
one of the categories would not engage in behaviors in
the other category in the pair.
Participants were given a description of the behaviors
that would be included in each category and were asked
to rate the co-occurrence of the pairs of categories on a
seven point scale with 1 being anchored with `Very
Unlikely to Co-Occur' and 7 being anchored with `Very
Likely to Co-Occur.' They were asked to make their
ratings with regard to employees in general rather than in
reference to a specific organization or workplace.
The co-occurrence of the categories was rated rather
than rating the co-occurrence of the 66 individual
counterproductive work behaviors used in the self-rating
task in Study 1. With 66 items, 2,145 pairs of items would
need to be rated; with 11 categories, the number of pairs
reduces to a more manageable 55. An incomplete design
International Journal of Selection and Assessment
MELISSA L. GRUYS AND PAUL R. SACKETT
could not be used, as the mail survey would not ensure an
adequate sample size for any given pair of items.
Analysis
Multidimensional scaling analysis using ALSCAL was
conducted using the data from the co-occurrence ratings.
Participants rated all possible pairs of CWB categories.
The co-occurrence ratings were used to directly form a
similarity matrix for use in the multidimensional scaling
analysis.
Results
Table 4 presents the similarity matrix. All of the mean
similarities between the pairs of items were in a fairly
small range, from 3.26 to 5.45. This indicates that, in
general, participants view all of the 11 CWB categories as
co-occurring to a certain degree. However, the
similarities suggest that participants do not view any of
the categories as extremely likely to co-occur, which
would be reflected by mean similarities closer to seven.
As input to the multidimensional scaling program, the
similarity matrix presented in Table 4 was converted into
a dissimilarity matrix by subtracting each of the
similarities from a constant. ALSCAL was used to
produce solutions for one to five dimensions. A scree
test was conducted by plotting the stress indexes for all
five configurations. The appropriate number of
dimensions can be determined on the basis of the scree
plot and where the stress index levels off. The one-
dimensional solution had a stress level of .34. The stress
level dropped to .12 for the two-dimensional solution,
and to .05, .03, and .01 for the three- four- and five-
dimensional solutions, respectively. The scree results
suggest either a two-dimensional or three-dimensional
solution. For the sake of interpretability and parsimony,
a two-dimensional solution was chosen. This solution is
presented in Figure 1.
Dimension 1 can be interpreted as reflecting an
Interpersonal versus Organizational dimension, or the
degree to which the behaviors are directed at an
individual or at the company. This dimension mirrors
the Interpersonal-Organizational dimension proposed by
the Robinson and Bennett (1995) typology. The behavior
categories that fall in the most extreme negative area on
this dimension contain mainly actions that are aimed at
individuals. Inappropriate physical actions (labeled
Physical in Figure 1), inappropriate verbal actions
(labeled Verbal), and behaviors which endanger others
due to lack of following safety procedures (labeled Safety)
are all Interpersonal acts in that they are aimed at
individuals. The behaviors within the categories of
Alcohol Use and Drug Use are not primarily aimed at
other individuals and are also not necessarily aimed at the
Blackwell Publishing Ltd 2003
DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR 39

Table 4: Similarity matrix for co-occurence data

Category 1 2 3 4 5 6 7 8 9 10

1. Theft and Related Behavior

2. Destruction of Property 4.25
3. Misuse of Information 4.17 4.08
4. Misuse of Time and 4.72 4.41 4.88
Resources
5. Unsafe Behavior 3.54 4.47 3.53 3.51
6. Poor Attendance 4.20 4.04 3.75 4.69 3.52
7. Poor Quality Work 4.20 4.51 4.11 4.65 4.24 5.45
8. Alcohol Use 4.28 4.59 3.94 4.26 4.77 5.24 5.31
9. Drug Use 4.50 4.64 3.97 4.28 4.78 5.26 5.29 5.13
10. Inappropriate Verbal 3.67 4.30 3.63 3.34 3.67 3.45 3.89 5.13 5.13
Actions
11. Inappropriate Physical 3.50 4.36 3.26 3.12 3.72 3.30 3.69 5.02 5.06 4.99
Actions

Note: Ratings were made on a 7-point scale with 7 being the most likely to co-occur.

company, which is reflected by their somewhat neutral
location on this dimension. The behavior categories that
fall in the most extreme positive area on this dimension
contain mainly actions that are aimed at the organization
rather than at individuals and are more impersonal
actions. Poor Attendance (Attend), Misuse of Time and
Resources (Timereso), and Misuse of Information
(Informat) are all categories that contain such behaviors.
Theft and Related Behavior also falls in the positive area
for the Interpersonal-Organizational dimension.
Dimension 2 can be interpreted as reflecting Task
Relevance. All of the behavior categories that fall on the
positive area of this dimension can be described as being
relevant to tasks that are performed within the context of
a job. In general, employees should attend work (Attend),
use time and resources appropriately (Timereso), have
high quality work (Workqual), and not engage in
activities that would put themselves or others in danger
(Safety, Alcohol and Drugs). The behavior categories in
the negative area reflect activities that are not related to
tasks performed within the context of a job. Engaging in
inappropriate physical or verbal actions toward others
(Physical and Verbal), theft and related behavior (Theft),
misuse of property (Property), and misuse of information
(Informat) can all be viewed as somewhat separate from
specific work tasks that must be performed within jobs.

Figure 1. Multi-dimensional scaling analysis solution for the co-occurrence rating data

Blackwell Publishing Ltd 2003 Volume 11 Number 1 March 2003

40
Discussion
This research yields a number of interesting results.
Correlational analyses of self-ratings in Study 1 indicate a
general pattern of positive correlations between all CWB
items. This indicates that as the likelihood that an
individual will engage in one type of CWB increases, the
likelihood of the individual to engage in another type of
CWB also increases.
Empirical support was found in Study 1 for the use of
categories of behaviors that were formed on the basis of
content themes. Factor analyses suggested a moderate fit
for an 11-factor model that corresponds to the 11 CWB
categories. Support for the use of the categories is useful
in that future studies can characterize counterproductive
behavior that occurs in the workplace in terms of these
content categories. The categories can be used as a means
of organizing extensive lists of behaviors. Further,
exploring the relationships between these categories will
continue to contribute to the formation of a taxonomy of
CWB. This, in turn, will contribute to the formation of
an overall theory of CWB.
In general, most of the background and demographic
variables were relatively uncorrelated with the categories
of CWB. However, age and work experience were
significantly negatively related to the categories. This
negative correlation indicates that older individuals are
less likely to engage in these types of CWB. In general,
this result is not unexpected given the results of past
research that suggests that older individuals are less likely
to engage in counterproductive behavior in the work-
place (e.g., Hollinger and Clark, 1983b). Many of the
past studies which found this result were conducted on
samples which were considerably younger and had
considerably less range in age than the present sample
and many were conducted using samples from retail
organizations (e.g., fast food restaurants) which employ
many young people. Therefore, it is notable that the
negative relationship between age and CWB still seems to
hold in this sample where the mean age is 40 years old, as
well. Work experience has a similar negative relationship
with the CWB categories.
The use of a sample of older, well-educated
individuals who hold primarily managerial and technical
level positions can be viewed as a limitation of the study
in the sense that this sample may not represent the
general working population well. It is encouraging,
though, that the results using our pilot sample of very
different individuals, namely, undergraduate students,
seem to converge for the self-ratings. However, the
student sample may also be viewed as unrepresentative of
the working population. Regardless of the representative-
ness of either sample, it is encouraging to see the results
using the self-ratings for both samples converge. Also,
since the study was conducted using an American sample,
the results might not generalize to the broader
International Journal of Selection and Assessment
MELISSA L. GRUYS AND PAUL R. SACKETT
international context. Along these lines, Murphy (1993)
mentions several cross-cultural differences in the
interpretations of possible counter-productive behaviors.
The two studies both produced a matrix indicating the
interrelationship among the same 11 categories of CWB.
Each study provides insight into the dimensionality of
CWB in a different way. In the case of the Study 1 self-
ratings, these interrelationships were indexed by
correlations. In the case of the Study 2 co-occurrence
ratings, these interrelationships were indexed by direct
judgments of the likelihood of co-occurrence, and the
similarity matrix was subjected to multidimensional
scaling. The two methods produced somewhat differing
pictures of the dimensionality of CWB.
The findings from the self-report data in Study 1 reveal
a strong association between all behavior items and
between all behavior categories. The correlations
between the CWB categories were all positive and ranged
from .17 to .71 and the average correlation was .43. This
is similar to the correlation of .46 (uncorrected) between
organizational and interpersonal deviance found by
Bennett and Robinson (2000). Results also reveal a
strong common underlying factor. This would suggest
that as the likelihood of an individual engaging in one
type of CWB increases, the likelihood of that individual
engaging in other types of CWB also increases. There are,
of course, differences in base rates across behaviors, so
this is not to suggest that an individual who will engage
in one of the counterproductive behaviors will also
engage in all other counterproductive behaviors.
The findings from the co-occurrence data in Study 2
indicate a two-dimensional solution. The first dimension
is an Interpersonal-Organizational dimension that
reflects whether the categories include behaviors that
are aimed at individuals or at the company. This
dimension parallels Robinson and Bennett's (1995)
Interpersonal-Organizational dimension. The second
dimension reflects Task Relevance, or the extent to
which the categories include behaviors that are related to
work tasks that are carried out on the job. One extreme
on this dimension includes the Misuse of Time and
Resources and Poor Work Quality categories, which
contain behaviors which are directly related to carrying
out work tasks. The other extreme on this dimension
includes the Inappropriate Physical Actions,
Inappropriate Verbal Actions, Theft and Related
Behavior, and Destruction of Property categories, which
contain behaviors that are not related to work tasks.
This two-dimensional solution differs from the two-
dimensional solution found by Robinson and Bennett
(1995). One dimension, the Interpersonal-Organizational
dimension found in this study corresponds to the
Interpersonal-Organizational dimension found by
Robinson and Bennett, but the other does not. Robinson
and Bennett's second factor was a Minor-Serious
dimension, whereas the second dimension in this study
Blackwell Publishing Ltd 2003
DIMENSIONALITY OF COUNTERPRODUCTIVE WORK BEHAVIOR
is Task Relevance. It appears that when asked to make
judgments of similarity without specifying further the
definition of `similarity', respondents attend to a Minor-
Serious dimension more strongly than in the case of the
present study, where similarity is explicitly defined in
terms of co-occurrence.
The difference in study findings between the Study 1
self-report data and the Study 2 co-occurrence data (as
well as the differences from other research findings, such
as those of Robinson and Bennett) indicate the need for
care in drawing conclusions about the dimensionality of
CWB. We had hoped that the two methodologies
employed here would converge on a common set of
findings. They do to some degree: in both studies what
emerges is a picture of positive relationships among the
various categories of CWB. Study 1, though, suggests a
single common factor, while Study 2 differentiates among
the CWB categories on two dimensions. We note that the
two methodologies do ask participants to take differing
perspectives: Study 1 asks for self-ratings, while Study 2
asks for participants' perceptions about whether
counterproductive behaviors co-occur among workers
generally and that perceptions of self and perceptions of
others may differ. Thus, continued careful work with
multiple methods is encouraged to further clarify the
structure of CWB.
A contribution of the study rests in the adoption of a
focus on the co-occurrence of behaviors. Co-occurrence
is used as a tool to examine the dimensionality of CWB.
While the two data collection strategies operationalize
co-occurrence differently (i.e., as correlations between
self-reports vs as direct judgments of likelihood of co-
occurrence), both operationalize `similarity' of behaviors
in terms of co-occurrence. In past literature, there had
been a focus on the `similarity' of behaviors rather than
on whether behaviors are likely to co-occur. Also, in
contrast to many past studies that address only one type
of counterproductive behavior, the current study
addresses the relationships between various types of
CWB. Determining the level of co-occurrence of
behaviors in the workplace is very informative to
organizations. The prediction and prevention of CWB
in organizations will be enhanced as the level of co-
occurrence of the behaviors becomes known. For
example, if there is research to suggest that certain
variables are significant predictors of one type of CWB
(e.g., theft) and that behavior is known to co-occur with
another type of CWB (e.g., destruction of property), the
predictors of theft may well be useful in predicting
destruction of property also. Am analogous situation
exists for the prevention of CWB.
Robinson and Bennett (1995), Robinson and
Greenberg (1998), Sackett and Devore (2001), and others
have acknowledged the need for the development of a
comprehensive theory of CWB that includes all of the
different types of CWB that can occur in the workplace.
Blackwell Publishing Ltd 2003
41
This study represents an important step toward a better
understanding of the structure or dimensionality of
CWB. Examining the relationships between various
counterproductive behaviors will eventually contribute
to more effective prediction and prevention of CWB in
the workplace.
Acknowledgements
This article is based on the thesis submitted by Melissa L.
Gruys as part of the requirements for her doctoral degree
in Human Resources and Industrial Relations at the
University of Minnesota. An earlier version of this article
was presented at the 16th Annual Meeting of the Society
for Industrial and Organizational Psychology, April
2001, San Diego, CA.
We gratefully acknowledge funding for this study by the
Department of the United States Navy, Office of Naval
Research grant N00014-98-1-0872 through the Personnel
Security Research Dissertation Award Program. The
interpretations, conclusions, and recommendations,
however, are those of the authors and do not necessarily
represent the position or policy of the agency.
Notes
1. Since t-tests indicated no mean differences on relevant
variables between participants who were not employed,
those who were employed part-time, and those who were
employed full-time, all were retained in the sample.
2. The articles from which items were obtained are listed with
an asterisk in the reference list.
References
Altheide, D.L., Adler, P.A., Adler, P. and Altheide, D.A. (1978)
The social meanings of employee theft. In J.M. Johnson and
J.D. Douglas (eds), Crime at the Top (pp. 90124).
Philadelphia: Lippincott.
*Baron, R.A. and Neuman, J.H. (1996) Workplace violence and
workplace aggression: Evidence on their relative frequency
and potential causes. Aggressive Behavior, 22, 161173.
Bennett, R.J. and Robinson, S.L. (2000) Development of a
measure of workplace deviance. Journal of Applied
Psychology, 85, 349360.
Bensman, J. and Gerver, J. (1963) Crime and punishment in the
factory: The function of deviancy in maintaining the social
system. American Sociological Review, 28, 588598.
Campbell, J.P. (1990) Modeling the performance prediction
problem in industrial and organizational psychology. In
M.D. Dunnette and L.M. Hough (eds), Handbook of
Industrial and Organizational Psychology (Vol. 1, pp. 687
732). Palo Alto, CA: Consulting Psychologists Press.
Cressey, D. (1953) Other People's Money: A Study in the Social
Psychology of Embezzlement. Belmont, CA: Wadsworth.
Fishbein, M. (1980) A theory of reasoned action: Some
applications and implications. In H. Howe and M. Page
Volume 11 Number 1 March 2003
42
(eds), Nebraska Symposium on Motivation, 27, 65116.
Fishbein, M. and Azjen, I. (1975) Belief, Attitude, Intention, and
Behavior: An Introduction to Theory and Research.
Reading, MA: Addison-Wesley.
Gouldner, A.W. (1954) Wildcat Strike: A Study in Worker-
Management Relationships. New York: Harper and Row.
Henry, S. (1978a) The Hidden Economy: The Context and
Control of Borderline Crime. London: Martin Robertson.
Henry, S. (1978b) Crime at work: The social construction of
amateur property theft. Sociology, 12, 245263.
*Hollinger, R.C. (1986) Acts against the workplace: social
bonding and employee deviance. Deviant Behavior, 7, 5375.
*Hollinger, R.C. and Clark, J.P. (1982) Formal and informal
social controls of employee deviance. The Sociology
Quarterly, 23, 333343.
*Hollinger, R.C. and Clark, J.P. (1983a) Deterrence in the
workplace: Perceived certainty, perceived severity and
employee theft. Social Forces, 62, 398418.
Hollinger, R.C. and Clark, J.P. (1983b) Theft by Employees.
Lexington, MA: Heath.
*Hollinger, R.C., Slora, K.B. and Terris, W. (1992) Deviance in
the fast-food restaurant: Correlates of employee theft,
altruism, and counterproductivity. Deviant Behavior: An
Interdisciplinary Journal, 13, 155184.
Horning, D.N.M. (1970) Blue collar theft: Conceptions of
property, attitudes toward pilfering, and work group norms
in a modern industrial plant. In E.O. Smigel and H.L. Ross
(eds), Crimes against Bureaucracy (pp. 4664). New York:
Van Nostrand Reinhold.
*Hunt, S.T. (1996) Generic work behavior: An investigation
into the dimensions of entry-level, hourly job performance.
Personnel Psychology, 49, 5183.
*Jones, J.W. (1980) Attitudinal correlates of employees'
deviance: Theft, alcohol use, and nonprescribed drug use.
Psychological Reports, 47, 7177.
*Mangione, T.W. and Quinn, R.P. (1975) Job satisfaction,
counter-productive behavior and drug use at work. Journal
of Applied Psychology, 60, 114116.
Mars, G. (1973) Chance, punters and the fiddle: Institu-
tionalized pilferage in a hotel dining room. In M. Warner
(ed.), The Sociology of the Workplace (pp. 200210). New
York: Halsted Press.
Murphy, K.R. (1993) Honesty in the Workplace. Pacific Grove,
CA: Brooks/Cole.
Ones, D.S. and Viswesvaran, C. (1998) Integrity testing in
organizations. In R.W. Griffin, A. O'Leary-Kelly and J.M.
Collins (eds), Dysfunctional Behaviors in Organizations,
Vol. 2. Nonviolent Behaviors in Organizations. Greenwich,
CT: JAI Press.
Ones, D.S. and Viswesvaran, C. (2001) Integrity tests and other
criterion-focused occupational personality scales (COPS)
used in personnel selection. International Journal of
Selection and Assessment, 9, 3139.
Ones, D.S., Viswesvaran, C. and Schmidt, F.L. (1993) Compre-
hensive meta-analysis of integrity test validities: Findings
and implications for personnel selection and theories of job
performance. Journal of Applied Psychology, 78, 4, 679703.
*Raelin, J.A. (1994) Three scales of professional deviance
within organizations. Journal of Organizational Behavior,
15, 483501.
International Journal of Selection and Assessment
MELISSA L. GRUYS AND PAUL R. SACKETT
Rise, J., Astrom, A.N. and Sutton, S. (1998) Predicting inten-
tions and use of dental floss among adolescents: An applica-
tion of the theory of planned behavior. Psychology and
Health, 13, 223236.
Robin, G.D. (1969) Employees as offenders. Journal of
Research in Crime and Delinquency, 6, 1733.
Robin, G.D. (1970) The corporate and judicial disposition of
employee thieves. In E.O. Simgel and H.L. Ross (eds),
Crimes against Bureaucracy (pp. 119142). New York: Van
Nostrand Reinhold.
*Robinson, S.L. and Bennett, R.J. (1995) A typology of deviant
workplace behaviors: A multidimensional scaling study.
Academy of Management Journal, 38, 555572.
Robinson, S.L. and Greenberg, J. (1998) Employees behaving
badly: Dimensions, determinants, and dilemmas in the study
of workplace deviance. In D.M. Rousseau and C. Cooper
(eds), Trends in Organizational Behavior (Vol. 5, pp. 130).
New York: Wiley.
*Robinson, S.L. and O'Leary-Kelly, A. (1996) Monkey see,
monkey do: The role of role models in predicting workplace
aggression. In J.B. Keys and L.N. Dosier (eds), Academy of
Management Best Paper Proceedings (pp. 288292).
Pleasantville, NY: Academy of Management.
Roy, D. (1952) Quota restrictions and goldbricking in a
machine shop. American Journal of Sociology, 57, 427442.
Roy, D. (1959) Banana time: Job satisfaction and informal
interaction. Human Organization, 18, 158168.
*Ruggiero, M., Greenberger, E. and Steinberg, L.D. (1982)
Occupational deviance among adolescent workers. Youth
and Society, 13, 423448.
Sackett, P.R. and Decker, P.J. (1979) Detection of deception in
the employment context: A review and critical analysis.
Personnel Psychology, 32, 487506.
Sackett, P.R. and DeVore, C.J. (2001) Counterproductive
behaviors at work. In N. Anderson, D.S. Ones, H.K.
Sinangil and V. Viswesvaran (eds), International Handbook
of Work Psychology (Vol. 1, pp. 145164). London: Sage.
Sackett, P.R. and Harris, M.M. (1984) Honesty testing for
personnel selection: A review and critique. Personnel
Psychology, 37, 221245.
Sackett, P.R. and Wanek, J.E. (1996) New developments in the
use of measures of honesty, integrity, conscientiousness,
dependability, trustworthiness, and reliability for personnel
selection. Personnel Psychology, 49, 787829.
Sackett, P.R., Burris, L.R. and Callahan, C. (1989) Integrity
testing for personnel selection: An update. Personnel
Psychology, 42, 491529.
*Skarlicki, D.P. and Folger, R. (1997) Retaliation in the
workplace: The roles of distributive, procedural, and inter-
actional justice. Journal of Applied Psychology, 82, 434443.
*Slora, K.B. (1989) An empirical approach to determining
employee deviance base rates. Journal of Business and
Psychology, 4, 2, 199218.
Taylor, L. and Walton, P. (1971) Industrial sabotage: Motives
and meanings. In S. Cohen (ed.), Images of Deviance (pp.
219245). London: Penguin.
*Terris, W. and Jones, J. (1982) Psychological factors related to
employees' theft in the convenience store industry. Psycho-
logical Reports, 51, 12191238.
Blackwell Publishing Ltd 2003