Human Resource Management Review 20 (2010) 6272

Contents lists available at ScienceDirect

Human Resource Management Review

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / h u m r e s

Reducing counterproductive work behavior through employee selection

Charles N. MacLane a,, Philip T. Walmsley b
a b

CorCompetencies, 2506 N. Upshur St., Arlington, VA 22207, USA U.S. Ofce of Personnel Management, USA

a r t i c l e

i n f o

a b s t r a c t
Counterproductive work behavior (CWB) has been a problem since organizations have hired employees. Recently, there has been increasing interest in explaining and addressing deviant behavior in the workplace. Our review of the research on CWBs shows a gap between theory development and the development of procedures to address deviant behavior. Moreover, studies in, for example, clinical psychology, have relevance for understanding CWBs, but crossfertilization with other relevant literatures is not in evidence. We summarize the contributions of three relatively distinct lines of research. We contend that current dimensional personality theories should guide the implementation of interventions seeking to reduce CWBs. We describe validity studies that incorporate the development of employee selection procedures based on the assumption that CWBs should be anchored within a dimensional model and we present empirical results that support the utility of that model. 2009 Elsevier Inc. All rights reserved.

Keywords: Personnel selection Counterproductive work behavior Workplace deviance Biographical data Personality Job performance Testing

Counterproductive work behavior (CWB) has been a problem since humankind has worked together to achieve communal goals. A papyrus recovered from an Egyptian tomb describes how, over 2000 years ago, workers building monuments for the nobility walked off the job to protest their lack of pay (Vernus, 2003). Of course, at that time and in that culture, strikes were a totally illegitimate and illicit activity. Although many different terms are found in the literature for CWBs workplace deviance, organizational misbehavior, and many others, they all describe behavior that is (a) negative for the current functioning of an organization and (b) is intentional (e.g., Grifn & Lopez, 2005; Sackett, 2002). CWBs encompass an extensive range of behaviors. Gruys' dissertation (as cited in Sackett, 2002) listed 87 different CWBs which could be grouped into 11 categories. Some of these categories, such as theft of property, clearly refer to intentional acts. Other categories of behavior, such as wasting time, are negative but are more difcult to assume as voluntary. In the present paper, we will summarize the variety of studies that characterize current thinking about CWBs. Our summary indicates that theoretical understanding of CWBs has grown rapidly in recent years, but that theoretical understanding does not inform the development of procedures for addressing CWBs, such as hiring workers who are less likely to engage in them. The latter have typically focused on preventing a specic set of behaviors, such as betraying company secrets or employee theft. Thus, their development has been largely atheoretical (or the theory has been left unexplained in the literature). We will describe an approach in which steps in the development of employee selection procedures explicitly build on theory about CWBs and are supported by empirical data. Further, we will present evidence that a model of CWBs that treats them as behaviors that are on continua with other job performance behaviors can be advantageous both in theory development and in the implementation of selection procedures.

Corresponding author. E-mail address: Chuck.MacLane@gmail.com (C.N. MacLane). 1053-4822/$ see front matter 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.hrmr.2009.05.001

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

63

1. Theories of counterproductive work behavior 1.1. Intentional counterproductive work behavior According to Grifn and Lopez (2005), the current ways of thinking about deviant behaviors in organizations have been formulated in the years since 1995. They argue that, during this period, there have been such a variety of conceptual approaches to understanding CWBs that this variety may actually be hindering progress. The multitude of behaviors identied in Gruys's dissertation (as cited in Sackett, 2002) catalogue of 87 CWBs and 11 behavioral categories indicate the difculty of deriving a single behavior-based approach to explaining deviant behavior in the workplace. In an additional study, Gruys and Sackett (2003) started with approximately 250 counterproductive work behaviors and reduced the list to 66 CWBs. Other researchers have continued to study CWBs but have worked to develop more parsimonious models. Among the earlier studies, Hollinger's (1986) division of deviant work behaviors into property and production deviance recognized distinctions between attacks on organizational processes and attacks on material resources of the organization, but did not recognize deviant behavior with respect to persons. In contrast, Robinson and Bennett (1995) identied deviant behavior among individuals (ID) as one of the two primary categories of such behavior. Their second category was deviance toward the organization (OD). Their self-report measure a workplace deviance scale (Bennett and Robinson, 2000) is based on this two-factor model. According to Berry, Ones, and Sackett (2007), it has been the most used measure of CWBs in studies of workplace deviance. However, Robinson and Bennett's (1995) emphasis on behaviors may have led to the divergence in the theoretical understanding of these categories of behaviors noted by Grifn and Lopez (2005). The Bennett and Robinson (2000) CWB measure consists of items that tap the frequency of the respondent's engagement in various CWBs. It was intended to be a survey reecting normative behavior, not a measure of constructs that explain deviant behavior. Bennett and Robinson (2000) did explore relationships between the two factors of their survey ID and OD and marker variables. These marker variables consisted of both measures of behavior, such as organizational citizenship behaviors (OCBs; Smith, Organ, & Near, 1983), and measures of attitudinal and motivational constructs, such as the Machiavellianism scale (Christie & Geis, 1970). The relationships were almost entirely consistent with expectations. A number of researchers have examined the relationships of cognitive, motivational, attitudinal, and dispositional constructs with CWBs. Dilchert, Ones, Davis, and Rostow (2007) found that cognitive ability predicted CWBs, with CWBs measured as records of formally recorded incidents. The researchers split the CWB criteria into categories based on Robinson and Bennett's ID/OD distinction. The validity coefcient for cognitive ability tests predicting overall CWBs was .33; when CWBs were split into dimensions, the validity coefcient was .36 for organizational CWB and .21 for interpersonal CWB. Dilchert et al. (2007) noted, individuals who were higher in cognitive ability engaged in CWB less frequently than did individuals who scored low on this trait (p. 622). At the same time, Dilchert et al. (2007) acknowledge that This raises the question whether the results obtained are solely due to the fact that more intelligent individuals get caught less frequently when engaging in CWB (p. 624). They further point out that the role of intelligence in criminal behavior is a point of controversy for those who study it. In defense of the validity of their use of organizational records as a criterion, Dilchert et al. (2007) cite studies showing that lack of intelligence is not correlated with being caught more, either in juveniles or adults. These studies must be characterized by the same conundrum, however, in that more intelligent wrong-doers could be able to escape identication more readily even in these studies and could even be less likely to report their own crimes in those studies where the criterion was self-reporting. At the present time, the Dilchert et al. (2007) study is the only published investigation linking cognitive ability with actual records of CWBs of which we are aware. Thus, much more research is needed before the role of intelligence in a theory of CWBs is established. Diefendorff and Mehta (2007) used latent variable modeling techniques to study the impact of motivational traits on CWBs. Using motivational trait constructs from Kanfer and Ackerman's (2000) research and Robinson and Bennett's (1995) ID/OD scales, they demonstrated relationships between approach and avoidance motivational traits and CWBs. Specically, they found that personal mastery (high value of learning and goal striving) was negatively related to ID and OD and BAS sensitivity (general sensitivity to rewards) was positively related to both ID and OD. Counter to expectations, competitive excellence (desire to compete or engage in social referencing) was unrelated to both dimensions of CWBs. Avoidance motivation was positively related to OD and was predictive of ID in the presence of organizational constraints. Hershcovis et al. (2007) conducted a meta-analysis on the predictors of workplace aggression, often dened as a component of the CWB domain. Using Robinson and Bennett's framework, they divided aggression into individual and organizational dimensions. They found that both individual difference and situational factors were related to aggression. Dalal (2005) meta-analytically investigated the antecedents of CWB and another job performance construct, organizational citizenship behavior (OCB; Organ, 1997). His results demonstrated that a number of motivational, attitudinal, and dispositional variables tend to relate to one's propensity to engage in CWBs. He found that job satisfaction, organizational commitment, and perceptions of organizational justice were negatively related to CWB. Regarding attitudinal and dispositional variables, Dalal (2005) found that both positive and negative affect were related to CWB, with negative affect displaying a slightly larger relationship to CWB. Finally, in accord with the observations of other researchers who have studied the relationship of various personality characteristics with CWB (Ones & Viswesvaran, 2003; Salgado, 2002; Sackett & DeVore, 2001), Dalal (2005) found a negative relationship between Conscientiousness and CWB. This nding was explored in further detail by Dudley, Orvis, Lebiecki, and Cortina (2006), who meta-analytically investigated the relationship of the narrow facets of Conscientiousness on CWB. Dudley et al. (2006) reported varying relationships of Conscientiousness subfacets with CWB, including dependability (r = .34), cautiousness (r = .11), and order (r = .07). Achievement was unrelated to CWB (r = .00). The lack of a relationship between achievement and CWBs in Dudley et al.'s meta-analysis is consistent with Diefendorff and Mehta's (2007) results.

64

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

Researchers focusing on contextual issues have linked variations in social relationships in the workplace to CWBs. Lawrence and Robinson (2007) demonstrated the relationship between power differentials within the work context and workplace deviance. Managers' inuence on employees' subsequent CWB was of particular importance in this study. They showed how interactions characterized by imbalances of power can lead to a loss of autonomy or perceived injustice in the workplace. Their ndings suggested that the loss of autonomy, or perceived injustice, creates frustration, which leads to resistance a type of CWB. Nevertheless, the researchers who studied the construct validity of CWB measures did not attempt to extend their analyses to address the issue of which of the constructs they studied were likely to be the proximal causes of the behavior and, as a consequence, they did not have a base from which to develop measures of constructs that could be used to predict a job applicant's tendency to behave counterproductively. 1.2. Employment tests There are multiple employment tests specically developed to predict which job candidates would be less likely to engage in CWBs. These operational tests were created to address the practical and urgent need of employers for information that could aid in hiring decisions. They share the traditional multiple-choice, self-report format. Of these tests, integrity or honesty tests (Camara & Schneider, 1994; Sackett, Burris, & Callahan, 1989; Ones & Viswesvaran, 2001a,b) predict the broadest range of CWBs. More narrowly-focused tests of this type include drug-and-alcohol tests, violence tests, and stress-tolerance tests (Ones & Viswesvaran, 2001a,b). In the 1980s, an estimated 6000 organizations administered up to 5,000,000 integrity tests annually (Sackett & Harris, 1984). Many of these tests were adopted by private sector businesses at the same time that restrictions on the use of the polygraph increased. It is particularly relevant that the development of these measures has rarely been reported in professional psychology journals and, as such, are not a part of the CWB literature that we have been reviewing. In fact, many integrity tests were based on polygraph examinations or were developed by polygraphers (Sackett et al., 1989). These considerations raised concerns about the technical adequacy and validity of these tests, much as they had with the polygraph (Camara & Schneider, 1994). Recently, some of these important technical issues were addressed by Ones and Viswesvaran (2003). They integrated and summarized a series of meta-analyses of validity studies for employment tests that appear to deal directly with counterproductivity: integrity/honesty tests, drug-and-alcohol tests, and violence tests. They labeled these tests criterion-focused occupational personality scales (COPS). Ones and Viswesvaran (2003) also conducted studies in which they administered these tests along with marker variables a cognitive ability test and measures of the Big Five personality variables: Conscientiousness, Openness, Stability, Extraversion, and Agreeableness (Costa & McCrae, 1992). Their meta-analyses showed that all these tests were substantially valid for criteria that consisted of various measures of CWBs, but they were equally or more valid for overall job performance criteria. Relevant to construct validity, it was found that Conscientiousness, Stability, and Agreeableness were the personality constructs that best explained variance with each type of counterproductivity test. In a theoretical framework, these results support the construct validity of these employment tests because these three personality constructs have particularly strong implications for how a worker relates to others, particularly in groups such as businesses. According to Digman (1997, pp12491250, cited in Ones and Viswesvaran, 2003), the three constructs comprise a higher order socialization factor that represents the socialization process itself. The empirical results of the Ones and Viswesvaran (2003) studies demonstrate that multiple-choice self-report measures can predict counterproductive work behaviors and that personality constructs can account for much of the variance in these measures. Nevertheless, Ones and Viswesvaran (2003) were not able to explicate a consistent theoretical basis for the development of these measures. Sackett et al. (1989) also noted this short-coming of these types of measures. It is paradoxical, then, that the CWB research literature (exemplied by studies such as Bennett and Robinson, 2000) has thoroughly investigated the constructs that were related to measures of behavior but the development of measures has apparently not been guided by theory based on these constructs. At the same time, those who have developed practical and useful measures have not been contributors to an understanding of why they work. The CWB measures employed by Bennett and Robinson (2000) and others were uniformly behavior-based and were capable only of predicting the counterproductive behavior of workers who had opportunities to engage in those behaviors in the past. Additionally, Ones and Viswesvaran (2001a,b) demonstrated that the integrity and other tests that they analyzed predicted overall job performance as well as specic CWB-related job criteria. This suggests that CWBs are not independent from other work behaviors in terms of the constructs with which they are related. 2. Clinical approaches to CWBs A third area of psychological research focuses on a different type of data and provides some important and complementary insights into the nature of CWBs. The traditional system of categories used by psychologists, psychiatrists, and other mental health professionals to dene mental illness begins by dividing all behavior into two major groups: normal and pathological. The pathological behaviors are themselves divided into a complex network of sub-categories that are dened essentially by differences in symptoms. This diagnostic system is codied in the Diagnostic and Statistical Manual of Mental Illness (DSM), currently in its fourth edition (American Psychiatric Association, 1994). The rst DSM was published in 1952 and has been the cardinal reference for legal and clinical purposes since then, but currently some researchers are questioning its credibility and usefulness. Widiger and Trull (2007) argue that present diagnostic DSM categories show excessive diagnostic comorbidity, inadequate coverage, arbitrary and unstable boundaries with normal psychological functioning, heterogeneity among persons sharing the same categorical diagnosis, and inadequate scientic base (p. 72). To address these problems, Widiger and Trull (2007) propose a

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272 Table 1 Illustration of personality continua encompassing both normal and abnormal behavior within a Big Five framework. Individuals who are too low on this dimension tend to be: Denitions of Big Five dimensions Conscientiousness Is degree of organization, motivation, and attention to details shown in work.

65

Individuals who are too high on this dimension tend to be: Obsessive, compulsive, rigidly selfdisciplined, may be workaholics

Careless, undisciplined; seen as disregarding rules and having trouble focusing on any standard for behavior

Openness to experience Narrow-minded and have little tolerance for different points of Is degree of curiosity, Openness to new experiences, Eccentric, impractical, constantly view or lifestyles willingness to consider unconventional ideas shown in work. changing goals, Agreeableness Is degree of concern with the needs of others, empathy, trust, Gullible, excessively generous, and pleasant disposition shown in work. often described as patsies Extraversion Is degree of outgoing behavior, warmth, eagerness to socialize shown in work.

Quarrelsome and suspicious; likely to manipulate others; seen as arrogant and lacking respect for law and society

Socially isolated; seen as lacking zest for life, shy, often described as not showing emotions

Seen as trying to dominate others, reckless, and demanding of constant attention

Consistently down, angry or anxious; impulsive in satisfying appetites (e.g., eating or sex); unable to control thoughts or feelings, often dependent on others

Adjustment Is degree of emotional Stability, secure and positive feelings, (No denition) and ability to act calmly and effectively under stress that is shown in work.

Note. This is the authors' graphic representation of the basic model presented in P. T. Costa and T. A. Widiger (1994). Personality disorders and the Five-Factor Model of Personality. Washington, D.C.: American Psychological Association.

fundamental shift in how psychopathology is conceptualized and diagnosed, more specically, a shift to a dimensional classication of personality disorder that would integrate the psychiatric classication with psychology's dimensional model of general personality structure (p. 71). Thus, Widiger and Trull (2007) would move beyond the dichotomous classication of normal and pathological behavior to a model that assumes pathological and normal behaviors lying on common dimensional continua. The dimensional model that they propose is, in fact, the Big Five personality framework (Costa & McCrae, 1992). Widiger and Trull's (2007) proposal is consistent with the empirical results reported by Ones and Viswesvaran (2001a,b) that integrity tests predict overall job behavior as well as they predict criteria specically related to CWBs. Both of these points support a conceptual approach that categorizes CWBs by the construct (that is, the dimension, characteristic, or trait) to which they are most closely related. We have developed a representation (Table 1) of the dimensional model of Widiger and Trull (2007) that highlights the types of behaviors that lie at the end of the Big Five dimensions. The characterizations of typical attitudes and behaviors lying at the negative ends of Conscientiousness, Agreeableness, and Stability match on a content basis with Gruys and Sackett's (2003) 11 categories of behaviors (Table 2): for example, theft and related behavior are most closely associated with a lack of Conscientiousness, destruction of property with low adjustment, and verbal and physical attacks on co-workers with low Agreeableness. All of the behavior categories appear, on a content basis, to reect one or more of these three dimensions. None of these behaviors have more than a trivial relationship to the remaining Big Five constructs: Openness and Extraversion. The associations among the categories t straightforwardly with the three personality characteristics Conscientiousness, Agreeableness, and Stability that Ones and Viswesvaran (2003) arrived at rather easily on a consistent basis.

Table 2 Behavior categories and examples of counterproductive work behaviors. Behavior category Theft and related behavior Destruction of property Misuse of information Misuse of time and resources Unsafe behavior Poor attendance Poor quality work Alcohol use Drug use Inappropriate verbal actions Inappropriate physical actions Note. Adapted from Gruys as reported by Sackett (2002). Example behaviors Theft of cash or property; giving away of goods or services; misuse of employee discount Deface, damage, or destroy property; sabotage production Reveal condential information; falsify records Waste time, alter time card, conduct personal business during work time Failure to follow safety procedures; failure to learn safety procedures Unexcused absence or tardiness; misuse sick leave Intentionally slow or sloppy work Alcohol use on the job; coming to work under the inuence of alcohol Possess, use, or sell drugs at work Argue with customers; verbally harass co-workers Physically attack co-workers; physical sexual advances toward coworker

66

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

Thus, the separation of CWBs from other job behaviors based primarily on the purposiveness of the behavior is also decient in the same way that the DSM dichotomy of normal versus pathological is: that is to say, the degree to which a behavior is purposive is itself on a continuum rather than being dichotomous. Furthermore, with respect to work behavior, there is clearly a measurement issue in determining where the behavior stands on a purposiveness continuum. For example, wasting time is considered a CWB if it is intentional, but it would be virtually impossible for an outside observer, or, indeed, often for the actor, to rate reliably the extent to which this common behavior is purposive. 3. CWBs in employee selection: Empirical evidence In this paper, we have reviewed several research literatures for their relevance to the prediction of deviant workplace behavior. Since the early 1900s, the Ofce of Personnel Management (the United States Civil Service Commission until 1980) has had responsibilities for the development and administration of many of the tests used to hire Federal government employees. In the past 15 years, there has been an accelerated focus on the noncognitive demands of occupations and, consequently, an increase in the development of selection procedures that measure noncognitive job competencies. We will present data showing that the incorporation of CWBs was intrinsic to the recent development of selection procedures that have become an important element in Federal government selection programs. 3.1. The Individual Achievement Record The rst entry level selection procedure that was specically developed to measure noncognitive work-critical competencies in the Federal government was a biodata instrument the Individual Achievement Record (IAR). The research that led to the IAR began in 1986 (Gandy, Dye, & MacLane, 1994). While evaluation of background information on the education and experience of job candidates had been a basic element of selection for the U. S. civil service during the most of the 20th century, the existing questions were based on content validity. Also, the scoring of this information was, for the most part, rationally derived (Lyons, 1989). However, most civil service jobs were entry-level and did not require specialized training and experience, thus limiting the utility of education and experience for predicting job performance. The other major selection tool used in hiring for the U. S. civil service was the written general mental ability (GMA) test. These tests were often empirically validated in criterion-related studies and were effective in predicting job performance, but it was clear that they did not address the important facets of job performance that were associated with noncognitive competencies. Moreover, they consistently showed high impact against minority groups. This led the Ofce of Personnel Management (USOPM), during the period 198689, to develop a test battery that included a biodata selection procedure the IAR that could be used for Federal government employee selection across agencies and occupations. This procedure was at once a continuation of the traditional civil service measurement of education and experience and at the same time a step forward in Federal government selection technology. The IAR was developed following a rational/empirical model that required items be both theoretically relevant and empirically valid (Gandy et al., 1994). The items selected for analysis in the validity study reected job analysis information on Federal professional and administrative occupations. Because the criterion was overall performance as rated by rst-line supervisors, the items included in the IAR reected noncognitive as well cognitive competencies. An extensive criterion-related validity study of approximately 100 professional and administrative occupations was conducted. From data on over 6000 government employees, the IAR was developed to select entry-level career workers who would be motivated and conscientious (Gandy et al., 1994). The approach that was formulated in the process of developing the IAR served as a guide for the later development of occupation-specic selection procedures. A particularly salient nding was that empirical results for biodata items varied considerably with the criteria against which they were validated (Gandy et al., 1994, pp. 303305). This result led to a focus in test development work on the meaning of the criterion as well as the theory and content of the test items and on collecting validity data with individual criterion elements e.g., specic job competencies as well as whole job measures. In particular, this focus led to the collection of critical incident (Flanagan, 1954) data. These data provided insight that was not accessible otherwise into work behaviors and the competencies on which they were based. In the present study, we analyze data that were collected in two different validity studies. These data were chosen for analysis because they demonstrate most effectively the role of CWBs in the measurement of noncognitive behavior. In these analyses, we sought to evaluate two hypotheses that owed from our literature review: Hypothesis 1. Critical incident behavioral data are valid bases for inferring the noncognitive competencies required for job performance and CWBs will constitute a substantial proportion of negative critical incidents. Hypothesis 2. Noncognitive job competencies based on critical incident data will explain and predict job performance. 3.2. Critical incident job analysis In one of the earliest projects (MacLane, 2006) in which development of a noncognitive entry-level selection procedure was based on a critical incident job analysis, incidents were collected from incumbent Federal employees who had extensive dealings with foreign nationals. Critical incidents (Flanagan, 1954) are descriptions of work behaviors that demonstrate good, poor, or

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

67

satisfactory performance. In form, they are short narratives that describe performance of one or more employees in a situation relevant to work. The outcome of the incident can be positive, negative, or neutral. Because they require specic details, critical incidents allow the linkage of behavioral content to job performance. We will provide evidence to show that an important subset of negative critical incidents are CWBs and that they have a substantial role in establishing the job constructs on which selection procedures are based. Two hundred thirty-one usable critical incidents were collected from incumbents in this study. The incident quoted below is a particularly striking example of one of the behaviors collected. The junior employee stationed in Guadalajara Mexico was scheduled to be on duty during an anti-U.S. demonstration. He was instructed to remain in the building, to observe and record events, and to advise superiors as necessary by telephone if anything extraordinary occurred. In the course of their protest, the demonstrators burned a paper U.S. ag and threw it over the fence into the yard. As the protest wound down and the demonstrators gave signs of departing, the enraged junior employee charged out of the building and threw the ag back over the fence. This act rekindled the demonstrators' ire, causing them to rush the fence. Police managed to restrain them. To determine whether the job competencies could be meaningfully determined from this type of data, the critical incidents were rated on the extent to which they demonstrated competencies. These ratings were factor analyzed. The analysis identied clear factors that were the basis for the competencies subsequently used to develop a selection procedure for the occupation. The results provided a positive basis for performing similar analyses with other critical incident data. The methodology of the rating process involved several steps. 3.2.1. Rating procedure Seventeen potential generalized dimensions of work were identied from previous job analyses. From the pool of 231 usable critical incidents, 100 were selected randomly. Six Ph.D. psychologists rated each incident on the degree to which the behavior of the protagonist(s) illustrated each of the seventeen job dimensions. The rationale supporting this rating process was that if expert judges who are familiar with the qualities that account for job performance consistently place two or more job performance dimensions together as causes of performance in a critical incident, then it can be inferred that those dimensions stand, to some meaningful degree, for the same underlying competency. The ratings were made on a ve-point scale ranging from Not at all to Almost completely. Thus, each rater made 1700 judgments evaluating each critical incident against each dimension. The six matrices (one per rater) were then merged into a single matrix by taking the average rating of each cell. The resulting 17 by 100 matrix of average ratings was factor analyzed. A maximum likelihood analysis with both oblique and orthogonal rotations was performed. In each rotation, seven solutions were specied: a solution for one factor, another for two factors, and so on up to seven. For each solution, and for each rotation, signicance tests, scree tests, and theoretical consistency were evaluated. The six factors shown in (Table 3) best represented the data in both oblique and orthogonal cases. These factors were easily interpretable as workrelated performance competencies. In sum, the critical incidents proved to be data from which job competencies could be educed. This important function was additional to the original purpose which they served: providing behavioral material from which test items could be developed. The critical incidents described actual work behaviors that dened positive or negative ways of performing, demonstrating clear

Table 3 Factor score coefcients displaying the relationship of derived factors on original dimensions in foreign national occupation study.

Note. Shading denotes factor score coefcients that are greater than the absolute value of 1.

68

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

relationships between work content and the constructs on which it was based. Thus critical incident analysis became an essential element of occupationally-specic test development. 3.2.2. Critical incidents dened as CWBs The factor analysis of ratings of critical incidents established that incidents t well within a theoretical framework of job-related competencies. We computed another factor analysis to investigate the relationships of the critical incidents that were judged to be CWBs. First, we computed a Q-type factor analysis that has been recommended as a superior technique for grouping jobs into families (Harvey, 1991). A Q-type factor analysis involves transposing the data matrix that is typically the subject of factor analysis so that the cases become variables and the variables become cases. Colihan and Burger (1995) conducted a Monte Carlo study to compare Q-type factor analysis and variations of it with the more commonly used hierarchical cluster analysis. The former techniques proved more robust under conditions of measurement error or overlap between job families; conditions that are typically found in the development of job families (Harvey, 1991). In the rst stage of this analysis, we transposed the original matrix of 17 generalized dimensions of work by 100 critical incidents so that the 17 dimensions became the cases in the factor analysis and the 100 critical incidents were analyzed as if they were the variables. We then computed the typical principal components analysis with varimax rotation. Based on the six factors that had been established in the initial factor analysis, we specied six factors in this factor analysis. The resulting analysis produced six groups of critical incidents which loaded on the six factors shown in Table 3. To clarify the meaning of these factors, we computed the factor score coefcient matrix (Table 3), which shows the coefcients by which variables are multiplied to obtain factor scores (note that the factors in this table are composed of combinations of critical incidents as shown in Table 4). We highlighted the factor scores in accordance with the strength of each factors' relationships with the competency dimensions: for example, the rst factor, Leadership, shows high, positive relationships with the Leadership and Self Direction dimensions. Overall, the meaning of each of the factors (that are composed of groups of critical incidents in this analysis) is clearly interpretable and consistent with the factors derived from the rst rating process. After the Q-type factor analysis provided us with the data showing the relationships of the critical incidents to meaningful factors, we rated these 100 critical incidents again to reveal which ones described counterproductive work behaviors. Two psychologists the rst and second authors of the present study independently rated each critical incident, rst to establish that the incident illustrated a negative behavior a behavior that illustrated poor performance, and second, to evaluate whether it was a CWB. Thirty-six incidents were rated as negative behaviors with complete agreement between the raters on the ratings of the

Table 4 Sample critical incidents representing counterproductive work behavior in the foreign national occupation. Factor Leadership and self direction Incidents rated as Example of critical incident CWBs 0 The section, originally consisting of four incumbents, was reduced to one incumbent after one person was evacuated to the States for medical reasons, another person was unexpectedly reassigned, and another departed on emergency visitation travel and did not return to post. The incumbent remaining in the section was unable to effectively prioritize any of her new duties and responsibilities. An incumbent consistently engaged individuals in heated political and policy debate, rather than completing his job based on law and regulation. The incumbent's behavior bogged down the work process in the busy section and raised potential security problems. The incumbent was often heard publicly disparaging a colleague's skill and intelligence. The supervisor one day heard the incumbent at the other end of the large ofce shouting and cursing loudly at his counterpart, who apparently in an effort to be helpful had brought him what turned out to be marginally useful information. The incumbent was required to prepare a country-specic brieng book for use by a traveling ofcial. The incumbent ignored the instructions provided by the supervisor and produced an unusable brieng package which the supervisor had to rewrite at the last minute. An incumbent was not familiar with the specic procedures for providing emergency nancial assistance. She nonetheless gave a citizen personal funds, and that individual departed the country. The incumbent later sought reimbursement but had obtained no signed receipt and in fact had no paperwork of any kind documenting the case. During the preparations for a visit from a former U.S. President, the incumbent provided some assistance and was assigned some tasks to perform during the visit itself. During the actual visit, however, he gave little attention to the tasks that he had been assigned. He devised countless ways to get close to the ex-president and his wife. The incumbent took an active and personal interest in an issue within her portfolio; she felt that her understanding of the issue was superior to Washington's knowledge of the issue. On one occasion she interpreted guidance in a very loose manner and prepared a brieng paper for her supervisor. The incumbent had put the supervisor (and the mission) in a position where they were acting in a manner which was not consistent with Washington's guidance. On one occasion, the incumbent submitted a telegram for clearance and a senior ofcial requested several changes. The incumbent contested the proposed changes but was told they were necessary. The drafting ofcer reluctantly made the senior ofcial's proposed changes but in doing so made some additional modications to the telegram's text which had the effect of negating the senior ofcial's basic substantive intent. The incumbent was assigned to make arrangements for a visiting Congressional Delegate. Although the incumbent scheduled some ofcial meetings, he failed to make any administrative arrangements. The incumbent had previously relied on other people in the ofce to do such work for him, but those other people had gotten tired of covering up for his repeated failures to plan and organize.

Social skill

Communication

Problem solving and 0 discretion Integrity 8

Planning

Note. If no incidents were rated as CWBs, the example represents negative performance but not a CWB.

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

69

one hundred incidents as positive or negative. In the second rating phase, the kappa statistic, calculated as an index of agreement beyond what could be expected by chance among the two raters, was = .60. The magnitude of such a kappa value indicates a substantial level of agreement among raters (e.g. von Eye & von Eye, 2008). Raters then discussed the critical incidents to reach consensus on those describing CWB. After discussion, we agreed that 13 critical incidents described behaviors that were counter to the mission of the organization and were intentional (Sackett, 2002). Thus, a substantial proportion of the negative critical incident behaviors, 36%, were CWBs. The incidents that were reliably rated as CWBs are highlighted in Table 4, which displays examples of the critical incidents that make up each of the factors. The notable result of this analysis is that 11 of the 13 CWBs that we identied are, as indicated in Table 4, related to integrity or social relationship constructs. This result converges with the Ones and Viswesvaran (2001b) conclusions that Conscientiousness and Agreeableness (along with Stability) are the most important constructs in explaining counterproductive behaviors. The results of these analyses support Hypothesis 1. They show that critical incident data can, when analyzed as behaviors that typically demonstrate specic competencies, be the basis for determining the important competencies to be measured in a selection procedure. (In the present study, the focus was on noncognitive competencies.) Although there is no test for statistically evaluating the proportion of critical incidents that were identied as CWBs because there is no a priori theory from which to build a model for that distribution, we concluded from a literature review that studies of CWBs have not considered them as being in the same competency domain as other work behaviors. Therefore, to show that 36% of a sample of negative behaviors were CWBs that could be classied as examples of Big Five personality dimensions and, particularly, that 85% of the CWBs were examples of Conscientiousness or Agreeableness, is substantial evidence of a dimensional view of CWBs. 3.3. CWBs and the federal attorney selection procedure A second validity study (MacLane & Cucina, 2007) that collected CWBs as critical incident data was to develop an entry-level selection procedure for applicants for attorney positions. The primary responsibility of these attorneys is to determine whether or not the applications of private citizens or corporations for registration of their businesses are consistent with applicable laws and regulations. The attorneys handle a docket of applications and interact with applicants and other external customers. According to the agency, the resource intensive nature of the selection process in use at the time of contact had hampered the agency's ability to reach out to a large and diverse applicant pool. The agency reviewed resumes and interviewed several hundred applicants each time that openings were announced. As a practical matter, the agency was forced to limit publicity of job announcements because of the logistical problems associated with assessing the large numbers of candidates that they could potentially attract. The agency requested an assessment procedure that could be administered online and would meet the requirements of the Uniform Guidelines on Employee Selection Procedures (1978). Given that the minimum qualications for this agency position required both a law degree and bar membership, most applicants to the position already possessed the critical cognitive and knowledge-based competencies for the position. Thus, the new assessment process would focus on noncognitive competencies. The biodata selection procedure labeled a Job Compatibility Assessment (JCA) that was subsequently developed and validated for this agency (MacLane & Cucina, 2007) is conceptually similar to the COPS tests reviewed by Ones and Viswesvaran (2001b) in that it is a multiple-choice occupational test that is criterion-focused. However, it is construct-oriented, targeting general noncognitive competencies rather than narrower behavioral areas such as drug abuse or employee theft. The JCA is a preemployment screening questionnaire that focuses on matching the competencies that job candidates possess with motivational, attitudinal, and personality-related competencies needed at work. In this validity study, critical incidents were a major source for item development and were also analyzed to dene the competencies that would be measured. To evaluate the role of CWBs in the selection procedure, we re-rated critical incidents collected in that study. Based on our ratings of another pool of critical incidents (MacLane, 2006) for CWB characteristics, we hypothesized that a subsample of critical incidents would be judged to be CWBs and we hypothesized that they would be found to be explained by, primarily, the competencies that Ones and Viswesvaran (2001b) identied: Conscientiousness, Agreeableness, and Stability. MacLane and Cucina (2007) used a clustering procedure to analyze the critical incidents collected in the study of attorneys and found six competencies that satisfactorily explained these behaviors. They developed a supervisory performance rating form based on the six competencies that was used as the criterion in the validity study of the JCA procedure. To assist the supervisors in their ratings, they inserted exemplar critical incidents in the supervisory form under the competency to which each incident was linked. These exemplar incidents were those which best represented the types of behaviors most illustrative of the competencies. There were 40 exemplar incidents that were negative (out of a total of 225 negative incidents) and therefore represented possible CWBs. Two psychologists the rst and second authors of the present study independently rated each of the 40 exemplar negative critical incidents and reliably ( = .45) classied 24 incidents as describing CWBs. Table 5 shows the allocation of positive and negative critical incidents and CWBs by competency and by facets of the overall competencies. The CWBs are an unexpectedly high proportion of the negative exemplar incidents, actually comprising the majority of negative incidents for Customer Service/ Teamwork and for Conscientiousness/Honesty. It should be noted that the original exemplar incidents were chosen without reference to the concept of CWBs and, in fact, that analysis was completed over three years ago. Table 6 provides examples of critical incidents rated to be CWBs for each facet of each of the competencies. The primary result to emerge from the categorization of critical incidents across CWBs is that almost all CWBs 23 of 24 were attributed to competencies that have been the primary ones with which they have been linked in the Ones and Viswesvaran (2001a) meta-analysis and in MacLane's (2006) study. Table 5 shows that 14 CWBs were classied with constructs that appear to be related to the Big Five trait of Agreeableness: that is, Customer Service/Teamwork, Courtesy/Agreeableness, and Condence/

70

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

Table 5 Linkage of CWBs and competencies in the attorney occupation study. Critical incidents (CIs) Competency/dimension name Customer service/teamwork Behavior relevant to other staff Behavior relevant to customers Conscientiousness/honesty Following rules, policy, and law Thoroughness; careful work Achievement Organization/attention to detail Courtesy/agreeableness Condence/sociability Condence Appropriate interpersonal activity Work compatibility N 111 68 55 45 8 41 72 29 6 13 Percentage 24% 15% 12% 10% 2% 9% 16% 6% 1% 3% Exemplar CIs Positive CIs 7 5 8 3 2 7 8 4 1 4 Negative CIs 7 5 8 3 2 4 8 5 3 3 Negative CIs rated as CWBs 5 3 7 1 1 1 4 2 0 0 CWBs as a percentage of negative CIs 71% 60% 88% 33% 50% 25% 50% 40% 0% 0%

Note. Percentages do not add up to 100% due to rounding.

Sociability. Nine CWBs were related to Conscientiousness/Honesty, which is certainly an alternate label for the Big Five Conscientiousness. The Big Five trait Stability does not appear to be related to any of these CWBs. Thus, as with the MacLane (2006) study, there is a convergence of these results with the results reported by Ones and Viswesvaran (2001a).

Table 6 Sample critical incidents representing counterproductive work behavior for each competency in the attorney occupation study. Competency/ dimension name Customer service/teamwork Behavior relevant to other staff Behavior relevant to customers Number of incidents describing CWB 5 3 Sample critical incident

An attorney constantly volunteers to do special projects and assignments to avoid sharing the workload of handling applications, the main job. The attorney received a telephone request from an outside attorney to amend a pending application. The attorney stated that he/she has no authority to approve the request and must check with a supervisor. In fact, the attorney did have the authority but did not want to be bothered.

Conscientiousness/honesty Following rules, policy, and law Thoroughness; careful work Achievement Organization/attention to detail

7 1

1 1

Courtesy/agreeableness

An attorney ran a publishing venture from his work computer and stored pornographic material on it. An attorney took a short cut in deciding a substantive issue in a case because it gave them increased work product and an easier letter to write. This caused the applicant to receive a second ofce action far in the future, after they had relied upon the previous determination. A newly-hired attorney told a mentor that they were willing to work only the minimum number of hours and were interested in only performing at the minimum level of production. An attorney ordered many cases at once, which needed to be completed within two weeks. The attorney looked through all of the cases but only worked on the ones that appeared easiest. This left a lot of problem cases on the attorney's docket. The attorney had to scramble to complete all these difcult cases in a short period of time. An attorney made an ugly comment when dealing with a temporary docket clerk, who had provided many months of devoted service to the law ofce. The attorney told the docket clerk that as the attorney was qthe one who went to law school,q and implied it was the docket clerk's responsibility to do clerical work.

Condence/sociability Condence

Appropriate interpersonal activity Work compatibility

A new attorney in a mentor arrangement with an experienced attorney would argue extensively with his mentor and refuse to admit that he might need to correct his work. This new attorney argued even when the mentor provided case law that expressly contradicted his position. An attorney spends too much time talking and socializing with others, either in person or on the phone. As a result, his production is low (at the lowest acceptable level) and he has lost out on a number of promotions he has sought. The attorney is unhappy that he is only focusing on one area of law. Attorney A has been examining applications for several years and feels that his legal skills are not being sharpened. He does not enjoy the repetitiveness of the job and therefore, does not produce quality work. Attorney A is unmotivated and does minimal work.

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

71

3.4. Criterion-related validity The study described in MacLane and Cucina (2007) outlines the steps that were taken to create a valid attorney JCA procedure. MacLane and Cucina (2007) described how CWBs collected as critical incidents became important in the development of the critical work competencies and in the building of a job performance criterion measure. We also alluded to the use of CWBs and other critical incidents in creating test items for the JCA procedure. Because critical incidents had a central place in the procedure's development, the criterion-related validity that is nally estimated for the procedure indirectly reects the validity of their use in creating it. In the MacLane and Cucina (2007) study, validity data were collected on a sample of 208 attorneys. Correlations between the incumbents' responses to the JCA items and the supervisory job performance ratings were subjected to crossvalidation to address capitalization on chance. The coefcients of three double-cross validities ranged from .33 to .47, with an average value of .38 (p = .01) uncorrected. This is a strong result, particularly for a cross-validated estimate that is an underestimate of the true uncorrected validity (Campbell, 1990) and it implies that process through which the JCA procedure was developed was likewise strong. Hypothesis 2 is supported by (a) the results that demonstrate that CWBs are exemplar behaviors for ve of six criterion performance rating scales and comprise the majority of the examplar behaviors in two of the scales and by (b) the signicant and substantial shrunken validities between these rating scales and the JCA biodata predictor questionnaire. 4. Discussion The current interest in CWBs has resulted in an increase in the number of studies seeking to understand these behaviors that cause so much loss, both in personnel and in resources. Our review indicated that research had proceeded on divergent paths: one primary direction had concentrated on investigating the constructs related to CWB (e.g., Dudley et al., 2006) while the other had been largely involved in the development of tests to identify job applicants who were likely to engage in various types of CWB (Ones & Viswesvaran, 2001b). We hypothesized that this separation might be partially caused by the difculty of measuring intentionality, the crucial variable differentiating CWBs from other negative work behaviors. We drew a parallel between the problems with this dichotomy and the separation in clinical research between normal and pathological behavior that is found in clinical research. The latter has also been strongly challenged as being inconsistent with data and theory. Finally, we drew on USOPM studies that report on the development of selection procedures in which critical incidents are important components and we collected data and performed analyses revealing that CWBs cannot be distinguished from other work behaviors in their measurement properties or construct relationships. We presented empirical data that supported our development of employee selection procedures that were specically built on this dimensional personality theory. We recognize that this paper has taken an explicitly individual differences viewpoint and has only in passing (Lawrence & Robinson, 2007) considered the contextual and organizational variables that have wide-ranging impact on employee behavior. It is also likely that other literatures in the social sciences have studied CWBs going from different assumptions. For example, we note an area of study focusing on ethical issues within organizations and among individuals in work settings. One paper (Andreoli & Lefkowitz, 2006) reported relationships between ethical climate in organizations and self-reported misconduct. The strongest relationship they observed was between the receipt of pressure from others in the organization to deviate from ethical guidelines and self-reported misconduct. Finally, it is clear from the evidence presented in this paper that there are specic areas in which future studies of CWBs could yield signicant dividends. First, there is a need for studies that establish measures of constructs that predict negative organizational behavior. Measures that predict behavior must go beyond counting frequency of past behaviors. They must tap personality variables, attitudes, and motivations. Such studies should focus on measures that have potential for predicting who is more likely to engage in purposively counterproductive work behaviors. In the studies we reviewed, there was an unexpected lack of relationships between measures of the Big Five trait Stability and behavioral indicants. Stability is the personal quality associated with lack of impulse control. One would expect that it would be strongly involved with CWBs. It is also clear that both individual difference and organizational/contextual variables explain differences in the frequency and type of CWBs that occur. As more construct-oriented measures are developed, studies that estimate the relative contributions of the various forces that effect CWBs can be planned. It is likely that the concept of CWB will eventually be resolved into a measure of tendencies to intentionally act out in work contexts. Acknowledgements We want to express our appreciation to Bernard Nickels and Patrick Sharpe for their helpful comments. The views expressed are those of the authors and do not necessarily reect the views of the Ofce of Personnel Management. References
American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders, 4th ed. Washington, DC: Author. Andreoli, N. & Lefkowitz, J. (2006). Individual and organizational antecedents of misconduct in organizations. Paper presented at the 13th Annual Conference of the Society for Industrial and Organizational Psychology, Dallas, TX. Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplace deviance. Journal of Applied Psychology, 85, 349360.

72

C.N. MacLane, P.T. Walmsley / Human Resource Management Review 20 (2010) 6272

Berry, C. M., Ones, D. S., & Sackett, P. R. (2007). Interpersonal deviance, organizational deviance, and their common correlates: A review and meta-analysis. Journal of Applied Psychology, 92, 410424. Camara, W. J., & Schneider, D. L. (1994). Integrity tests: Facts and unresolved issues. American Psychologist, 49, 112119. Campbell, J. P. (1990). Modeling the performance prediction problem in industrial and organizational psychology. In Dunnette, Marvin D. (Ed); Hough, Leaetta M. (Ed). Handbook of industrial and organizational psychology, Vol. 1 (2nd ed.). (pp. 687732). Palo Alto, CA, US: Consulting Psychologists Press. Christie, R., & Geis, F. L. (1970). Studies in Machiavellianism. San Diego, CA: Academic Press. Colihan, J., & Burger, G. K. (1995). Constructing job families: An analysis of quantitative techniques used for grouping jobs. Personnel Psychology, 48, 563586. Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO-PI-R) and NEO Five-Factor Inventory (NEO-FFI) Professional Manual. Odessa, FL: Psychological Assessment Resources. Costa, P. T., & Widiger, T. A. (1994). Personality disorders and the ve-factor model of personality. Washington, DC: American Psychological Association. Dalal, R. S. (2005). A meta-analysis of the relationship between organizational citizenship behavior and counterproductive work behavior. Journal of Applied Psychology, 90, 12411255. Diefendorff, J. M., & Mehta, K. (2007). The relations of motivational traits with workplace deviance. Journal of Applied Psychology, 92, 967977. Dilchert, S., Ones, D. S., Davis, R. D., & Rostow, C. D. (2007). Cognitive ability predicts objectively measured counterproductive work behaviors. Journal of Applied Psychology, 92, 616627. Digman, J. M. (1997). Higher order factors of the Big Five. Journal of Personality and Social Psychology, 73, 12461256. Dudley, N. M., Orvis, K. A., Lebiecki, J. E., & Cortina, J. M. (2006). A meta-analytic investigation of Conscientiousness in the prediction of job performance: Examining the intercorrelations and the incremental validity of narrow traits. Journal of Applied Psychology, 91, 4057. Flanagan, J. C. (1954). The critical incident technique. Psychological Bulletin, 51(4), 327358. Gandy, J. A., Dye, D. A., & MacLane, C. N. (1994). Federal government selection: The Individual Achievement Record. In G. S. Stokes, M. D. Mumford, & W. A. Owens (Eds.), Biodata handbook: Theory, research, and use of biographical information in selection and performance prediction (pp. 275310). Palo Alto, CA: CPP Books. Gruys, M. L., & Sackett, P. R. (2003). Investigating the dimensionality of counterproductive work behavior. International Journal of Selection and Assessment, 11, 3042. Grifn, R. W., & Lopez, Y. P. (2005). Bad behavior in organizations: A review and typology for future research. Journal of Management, 31, 9881005. Harvey, R. J. (1991). Job Analysis. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of Industrial and Organizational Psychology (pp. 71163)., 3rd ed. Palo Alto, CA: Consulting Psychologists Press, Inc. Vol. 2. Hershcovis, M. S., Turner, N., Barling, J., Arnold, K. A., Dupre, K. E., Inness, M., et al. (2007). Predicting workplace aggression: A meta-analysis. Journal of Applied Psychology, 92, 228238. Hollinger, R. C. (1986). cts against the workplace: Social bonding and employee deviance. Deviant Behavior, 7, 5375. Kanfer, R., & Ackerman, P. L. (2000). Individual differences in work motivation: Further explorations of a trait framework. Applied Psychology: An International Review, 49, 470482. Lawrence, T. B., & Robinson, S. L. (2007). Ain't misbehavin: Workplace deviance as organizational resistance. Journal of Management, 33, 378394. Lyons, T. J. (1989). Validity of education and experience measures in traditional rating schedule procedures: A review of the literature. Washington, DC: U.S. Ofce of Personnel Management, Competency Assessment Branch. MacLane, C. N. (2006). Alternative examination program of the State Department: Background and development. Washington, DC: Ofce of Personnel Management, Competency Assessment Branch. MacLane, C. N., & Cucina, J. M. (2007). Development and validation of a job compatibility assessment for trademark attorneys. Technical report Washington, DC: U.S. Ofce of Personnel Management, Competency Assessment Branch. Ones, D. S., & 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. (2001). Personality at work: Criterion-focused occupational personality scales used in personnel selection. In B. Roberts & R. Hogan (Eds.), Personality psychology in the workplaceDecade of Behavior. (pp. 6392) Washington, DC, US: American Psychological Association. Ones, D. S., & Viswesvaran, C. (2003). Personality and counterproductive work behaviors. In M. Koslowsky, S. Stashevsky, & A. Sagie (Eds.), Misbehavior and dysfunctional attitudes in organizations (pp. 211249). Hampshire, UK: Palgrave Macmillan. Organ, D. W. (1997). Organizational citizenship behavior: It's construct clean-up time. Human Performance, 10, 8597. Robinson, S. L., & Bennett, R. J. (1995). A typology of deviant workplace behaviors: A multidimensional scaling study. Academy of Management Journal, 38, 555572. Sackett, P. R. (2002). The structure of counterproductive work behaviors: Dimensionality and relationships with job performance. International Journal of Selection and Assessment, 10, 511. Sackett, P. R., Burris, L. R., & Callahan, C. (1989). Integrity testing for personnel selection: An update. Personnel Psychology, 42, 491529. Sackett, P. R., & Harris, M. M. (1984). Honesty testing for personnel selection: A review and critique. Personnel Psychology, 37, 221245. Sackett, P. R., & DeVore, C. J. (2001). Counterproductive behaviors at work. In N. Anderson, D. Ones, H. Sinangil, & C. Viswesvaran (Eds.), Handbook of Industrial, Work, and Organizational Psychology, Vol. 1 (pp. 145164). Thousand Oaks, CA: Sage Publications Ltd. Smith, C. A., Organ, D. W., & Near, J. P. (1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68, 653663. Salgado, J. F. (2002). The Big Five personality dimensions and counterproductive work behaviors. International Journal of Selection and Assessment, 10, 117125. Uniform Guidelines on Employee Selection Procedures. (1978). 43 Fed. Reg. 38, 29038, 315. Vernus, P. (2003). Affairs and scandals in Ancient Egypt. Ithica, NY: Cornell University Press. von Eye, A., & von Eye, M. (2008). On the marginal dependency of Cohen's . European Psychologist, 13, 305315. Widiger, T. A., & Trull, T. J. (2007). Plate tectonics in the classication of personality disorder: Shifting to a dimensional model. American Psychologist, 62, 7183.