Do Other-Reports of Counterproductive Work Behavior ... - CiteSeerX [PDF]

Journal of Applied Psychology 2012, Vol. 97, No. 3, 613– 636

© 2011 American Psychological Association 0021-9010/12/$12.00 DOI: 10.1037/a0026739

Do Other-Reports of Counterproductive Work Behavior Provide an Incremental Contribution Over Self-Reports? A Meta-Analytic Comparison Christopher M. Berry, Nichelle C. Carpenter, and Clare L. Barratt Texas A&M University Much of the recent research on counterproductive work behaviors (CWBs) has used multi-item selfreport measures of CWB. Because of concerns over self-report measurement, there have been recent calls to collect ratings of employees’ CWB from their supervisors or coworkers (i.e., other-raters) as alternatives or supplements to self-ratings. However, little is still known about the degree to which other-ratings of CWB capture unique and valid incremental variance beyond self-report CWB. The present meta-analysis investigates a number of key issues regarding the incremental contribution of other-reports of CWB. First, self- and other-ratings of CWB were moderately to strongly correlated with each other. Second, with some notable exceptions, self- and other-report CWB exhibited very similar patterns and magnitudes of relationships with a set of common correlates. Third, self-raters reported engaging in more CWB than other-raters reported them engaging in, suggesting other-ratings capture a narrower subset of CWBs. Fourth, other-report CWB generally accounted for little incremental variance in the common correlates beyond self-report CWB. Although many have viewed self-reports of CWB with skepticism, the results of this meta-analysis support their use in most CWB research as a viable alternative to other-reports. Keywords: counterproductive work behavior, workplace deviance, self-report, meta-analysis

Spector, & Miles, 2001), that include a wide range of CWBs (Rotundo & Spector, 2010). These multi-item measures of CWB acknowledge the positive manifold across individual CWBs, and the theoretical perspective that these individual behaviors reflect a broad, hierarchical construct with a general CWB factor at the top level, group factors such as interpersonal- and organizational target CWB beneath the general factor, and specific CWBs below the group factors (Berry, Ones, & Sackett, 2007; Sackett & DeVore, 2001). Furthermore, this broad CWB construct has been shown to relate to various organizationally relevant variables such as Big Five personality, negative affectivity, organizational justice, and job satisfaction, among others (Berry, Ones, & Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007). Thus, there is value in measuring and predicting a broad CWB domain (Sackett, 2002). The present meta-analysis concentrates on studies with such a focus on broad measurement of CWB via multi-item measures. A number of recent meta-analyses have used broad, multi-item CWB measures to address important research questions surrounding CWB. For instance, Berry, Ones, and Sackett’s (2007) metaanalysis supported Sackett and DeVore’s (2001) proposition that CWB is a hierarchical construct consisting of separate interpersonal- and organizational target lower order factors. Metaanalyses have also increased understanding of CWB’s nomological net by summarizing the relationships that CWB has with a wide range of organizationally relevant variables (e.g., Berry, Ones, and Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007). Although these meta-analyses have advanced knowledge related to the nature of CWB, these meta-analyses have in common one important limitation. Specifically, the primary studies included in each of the above meta-analyses relied almost completely on single-source

Counterproductive work behaviors (CWBs) refer to voluntary employee behaviors that are viewed by the organization as contrary to its legitimate interests, violate significant organizational norms, and threaten the well-being of the organization or its members (e.g., theft, withholding effort, interpersonal aggression, poor attendance; Bennett & Robinson, 2000; Sackett & DeVore, 2002). Studies have estimated that CWBs not only cost organizations billions of dollars annually (Camara & Schneider, 1994; Murphy, 1993; Vardi & Weitz, 2004), but they have negative consequences for employees as well. For instance, being the target of CWB can lead to decreased job satisfaction and increased stress and intentions to quit, among other things (Budd, Arvey, & Lawless, 1996; Glomb, 2002). As a result, a great deal of research has been devoted to the measurement and prediction of CWBs (Berry, Ones, & Sackett, 2007). Until recently, there was a tendency to treat each type of CWB as discrete, resulting in separate literatures focused on the measurement of specific CWBs such as theft or harassment (Sackett, 2002). Although separate literatures devoted to individual CWBs still exist and flourish, recent years have seen an increased interest in more integrative measurement of CWB. Much of the recent research on CWB uses multi-item measures, such as the Workplace Deviance Scale (Bennett & Robinson, 2000) and Counterproductive Work Behavior Checklist (Fox,

This article was published Online First December 26, 2011. Christopher M. Berry, Nichelle C. Carpenter, and Clare L. Barratt, Department of Psychology, Texas A&M University. Correspondence concerning this article should be addressed to Christopher M. Berry, Department of Psychology, Texas A&M University, 4235 TAMU, College Station, TX 77843-4235. E-mail: [email protected] 613

614

BERRY, CARPENTER, AND BARRATT

self-reports of CWB.1 This is not surprising, considering that the most popular multi-item CWB measures were originally designed as self-report measures (e.g., Bennett & Robinson, 2000; Fox et al., 2001). However, there have been recent calls to use such multi-item measures to collect supervisor and/or coworker ratings of CWB (we hereafter refer to supervisor and coworker ratings as other-ratings) as alternatives or supplements to self-ratings (e.g., Fox, Spector, Goh, & Bruursema, 2007; Stewart, Bing, Davison, Woehr, & McIntyre, 2009). This highlights two gaps in the present understanding of CWB. First, it is not clear whether the above meta-analyses’ results may have been affected by a range of issues germane to the use of self-report measures such as common method variance and social desirability bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). Second, it is unclear the extent to which the present understanding of CWB— obtained primarily from self-report measures— generalizes to CWB assessed through non-self-report measures. We address these gaps in the present study by empirically investigating a number of important issues regarding the degree to which using self- versus other-ratings makes a difference in CWB research. For instance, what is the correlation between self- and other-reports of CWB? Do self- and other-reports of CWB relate differently to theoretically important variables? Do self- and otherraters report different mean levels of CWB for employees? Does a pattern of results emerge suggesting one method of measurement is preferable? We use meta-analysis in the present study to draw on multiple samples and thousands of participants, which therefore represents a powerful test of the above research questions.

Review of the Literature and Development of Hypotheses Most fundamentally, we ask in the present study: Are self-report CWB measures a problem, and are other-reports the solution? Self-report has long been a common mode of measurement for CWB researchers (Fox et al., 2007). This is in part due to a number of important advantages that self-report measures of CWB have over other-report CWB measures. First, because many CWBs are relatively covert behaviors that employees engage in with the intention of not getting caught, the only source that has complete knowledge of an employee’s engagement in CWB is the employee. Second, having employees anonymously report their own CWB circumvents some ethical concerns associated with supervisor or coworker reports (Fox & Spector, 1999). That is, drawing attention to employees’ negative organizational behaviors by asking their supervisors or coworkers to rate the employees’ CWB could have negative consequences for the employees, calling into question whether risk is adequately minimized. Third, publicly available, multi-item, and validated self-report measures of CWB have existed for years (e.g., Bennett & Robinson, 2000; Fox & Spector, 1999; Marcus, Schuler, Quell, & Humpfner, 2002), and it is easier to administer such measures to employees themselves than to enlist supervisor or coworker raters. Regardless, self-report measures of CWB are not without their disadvantages. Unlike many variables in applied psychology, CWB involves highly sensitive inquiries about potentially selfincriminating information. So, one concern is the possibility that employees underreport the extent to which they engage in CWB (e.g., Fox et al., 2007). This underreporting might be due to the

fear of getting caught and being punished (R. M. Lee, 1993) or due to a general reluctance to describe oneself in negative terms (e.g., Heneman, Heneman, & Judge, 1997). Another concern with selfreport CWB is common method bias (e.g., Campbell & Fiske, 1959). When employees provide self-reports of CWB along with self-reports of other relevant variables, relationships between CWB and these other variables can be artificially inflated. These sorts of disadvantages have recently caused a number of researchers to voice concerns over the use of single-source selfreport measurement of CWB. For example, Barclay and Aquino (2010) listed self-report measurement as a methodological issue that has posed an obstacle for CWB research. Fox et al. (2007) made the point that although using self-reports from employees may be the most viable methodology for early stages of CWB research, “the CWB community is now at the point that more objective, or nonincumbent, measures are needed to further our understanding of the phenomenon” (p. 43). Other recent articles have made similar points (e.g., Stewart et al., 2009). Implicit in all of these concerns is the idea that the present CWB knowledge base derived using multi-item self-report measures is deficient and perhaps even misleading and that the use of other sources of CWB information will tell us something new about explaining and predicting CWB. Thus, CWB researchers have been relying more and more on other-reports of employees’ CWB, despite the difficulty of collecting such other-reports relative to collecting self-reports. Otherreports of CWB mitigate some of the key concerns over self-report CWB (e.g., common method bias, underreporting due to fear of being caught). However, there are clear disadvantages to otherreports of CWB. Supervisors and coworkers may not have adequate opportunity to observe employees engaging in CWB. Additionally, other-raters may fear retribution for reporting employees’ CWB and therefore may be reluctant to provide accurate information (Fox et al., 2007). So, if researchers are to go to all of the trouble of collecting other-ratings (or multisource ratings) of CWB, then it is important to understand what additional, unique information is gained beyond what is presently known about self-report CWB.

1 This is not meant to imply that virtually all past CWB research has relied on self-report CWB. For example, literatures devoted to specific CWBs, such as absenteeism or theft, have often used organizational records or other non-self-reports. Additionally, Ones, Viswesvaran, and Schmidt’s (1993) meta-analysis of the validity of integrity tests for predicting CWB carried out moderator analyses wherein self-report CWB measures were considered separately from other-report CWB measures (although many of the CWB measures were narrow, individual CWBs such as absenteeism and theft, in contrast to the broad, multi-item measures including a wide range of CWBs that are the focus of the present study). More recently, studies by Dilchert et al. (2007) and Stewart et al. (2009), among others, have used broad, multi-item, nonself-report measurement of CWB. We simply make the point in the present article that the vast majority of primary studies in the past decade in which broad, multi-item measures of CWB have been used have relied on self-report. Therefore, the cited meta-analyses that focused on prediction of broad, multi-item measures of CWB also were forced to rely mostly on single-source self-reports.

SELF- AND OTHER-REPORTS OF CWB

Relationship Between Self- and Other-Report CWB The correlation between self-reports and other-reports of CWB was meta-analyzed in the present study. If the correlation is high, then this suggests that self- and other-ratings of CWB provide little unique information. In this case, collecting CWB data from one source would be essentially equivalent to collecting it from the other source or from multiple sources, and therefore the efficacy of collecting data from sources other than the self would be called into question. If the correlation between self- and other-ratings of CWB is low, then this would suggest that the two sources provide highly unique information. To the degree that the unique information from each source accurately represents true levels of CWB, this means potential for incremental validity is high and that multisource ratings of CWB might increase knowledge and prediction of CWB. This would also suggest that the cumulative literature on CWB, which has relied mostly on self-report CWB (e.g., Berry, Ones, & Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007), is not generalizable beyond self-report. This would mean that new meta-analytic evidence regarding the nomological net of other-report CWB would be needed. There are reasons to expect that self-reports and other-reports of CWB will each provide unique information but still overlap to some degree. Some reasons to expect unique information are that other-raters may not have an opportunity to observe all of the CWBs that employees engage in and that self-raters may misreport CWBs for fear of being caught or other impression management concerns. It is, however, unlikely that other-raters will be completely unaware of employees’ engagement in CWB (especially relatively public CWBs such as harassment or gossip) and that the employees themselves will be completely untruthful about their engagement in CWB (if employees were completely untruthful and reported not engaging in any CWBs, it would be difficult to reconcile this with the cumulative literature showing that self-report CWB correlates with a number of relevant variables); thus, it is likely there will be some overlap in self- versus other-ratings. Therefore, similar to multisource ratings in other domains such as overall job performance (Conway & Huffcutt, 1997) and personality (Mount, Barrick, & Strauss, 1994), it is hypothesized that there will be a positive correlation between self- and other-ratings of CWB but that this correlation will not be perfect even when corrected for error of measurement. Specifically: Hypothesis 1: The correlation between self- and other-ratings of CWB will be positive and greater than zero but will not reach unity. There are reasons to expect certain variables to moderate the relationship between self- and other-reports of CWB. For one, the source of the other-rating may moderate this relationship. Specifically, employees probably make greater efforts to hide their CWB from their supervisors, who can formally punish them, than their coworkers. Additionally, in most settings, coworkers probably work in closer proximity on a day-to-day basis with employees than do supervisors, and accordingly, coworkers should have greater opportunity to observe and report employees engaging in CWB. Therefore:

615

Hypothesis 1a: The correlation between self-ratings and coworker ratings will be greater than the correlation between self-ratings and supervisor ratings of CWB. Another variable that is expected to moderate the relationship between self- and other-ratings of CWB is whether the CWBs are public versus private. As described earlier, the more public the CWBs, the more opportunity other-raters should have to observe those behaviors and make sound judgments of employees’ levels of CWB, and the more self- and other-ratings of CWB should converge. Interpersonal target CWBs (e.g., cursing or being rude to others) are generally more public behaviors than organizational target CWBs (e.g., theft of organizational property). Therefore: Hypothesis 1b: The correlation between self- and otherratings will be higher for interpersonal target CWB than for organizational target CWB. A third variable that is expected to moderate the relationship between self- and other-ratings of CWB is the degree to which assurances of anonymity have been given to raters. Both selfand other-raters are more likely to provide accurate/truthful ratings of employees’ CWB when they feel more confident that their responses are anonymous and confidential, and thus, their ratings should show more convergence. Although most CWB studies attempt at least some assurances of anonymity for participants, there are differences between studies in the strength of these assurances. For instance, some CWB studies may simply tell participants their responses are anonymous, whereas other studies may take precautions such as having participants mail their completed surveys directly to the researchers or assigning participants secret codes. The important point is whether participants perceive there is greater anonymity of their responses, regardless of whether these sorts of steps truly do assure greater anonymity. Therefore: Hypothesis 1c: The correlation between self- and otherratings of CWB will increase as researchers take more steps to assure participants of the anonymity of their responses.

Relationships With Theoretically Relevant Common Correlates Hypothesis 1 suggests that self- and other-report CWB will each reflect some unique variance. However, even if other-ratings capture unique variance, if the unique variance is not valid—meaning that the variance captured by other-rated CWB does not explain incrementally beyond that already explained by self-reported CWB—it may not be worth going to the trouble to collect otherratings of CWB. Thus, the second main purpose of the present meta-analysis was to advance understanding of the degree to which other-ratings of CWB have different patterns or magnitudes of relationships with a common set of theoretically relevant variables. If self- and other-ratings of CWB contain unique and valid information, they should exhibit different patterns and magnitudes of correlations with these theoretically relevant variables. Theoretical accounts have made the case that CWBs should be related to a number of important predictors and correlates, which has prompted meta-analyses of the relationships between CWB and

616

BERRY, CARPENTER, AND BARRATT

the Big Five personality traits (Berry, Ones, & Sackett, 2007), negative affectivity (Hershcovis et al., 2007), demographic variables (Berry, Ones, & Sackett, 2007), justice perceptions (Berry, Ones, & Sackett, 2007), interpersonal conflict (Hershcovis et al., 2007), organizational constraints (Herschcovis et al., 2007), job satisfaction (Dalal, 2005), and organizational citizenship behaviors (Berry, Ones, & Sackett, 2007). Via correspondence with the lead authors of each of the listed meta-analyses, it was determined that each of these meta-analyses relied almost solely on self-report multi-item measures of CWB. Therefore, in the present study, each of these relationships was meta-analyzed, but only including multiitem other-reports of CWB. A comparison of the meta-analytic estimates from the present meta-analysis versus the previous metaanalyses tests the degree to which self-ratings and other-ratings of CWB have similar versus dissimilar patterns of relationships with a common set of theoretically relevant correlates, and therefore the degree to which self- and other-ratings account for unique and valid variance in true CWB. However, if other-ratings have different patterns or magnitudes of relationships with common correlates, this could also be driven by problems in ratings from at least one of the sources. For instance, if relationships differ in magnitude (e.g., self-ratings are more strongly related to correlates than other-ratings), common method variance could be inflating relationships between variables. To the degree that this is true, the differing magnitudes of relationships that self- and other-report CWB have with their common correlates could signal a problem with self-report measures, instead of being evidence that self-reports capture unique valid variance. Thus, before one can conclude that differing patterns or magnitudes of relationships with common correlates are evidence of unique valid variance, one must test other plausible explanations for the unique relationships (e.g., common method variance). To this end, we propose two hypotheses (2a and 2b) in the present meta-analysis regarding reasonable alternative explanations for differing patterns/magnitude of relationships self- and other-reports have with their common correlates. First, if common method variance is an issue, then self-reports of CWB will correlate more strongly with the common correlates (or at least those common correlates measured via self-report) than will otherreports of CWB. Therefore: Hypothesis 2a: Self-ratings of CWB will correlate more strongly than other-ratings of CWB with a common set of theoretically relevant correlates. Self- and other-report CWB can also differ in the pattern of relationships they have with their common correlates (e.g., self- and other-report CWB may not be similarly related to the same correlates). A common concern is that many self-report measures are affected by social desirability concerns (e.g., Spector, 2006). To the degree that this concern is legitimate, self-report CWB measures would be expected to correlate more strongly with additional selfreported variables that are similarly affected by social desirability concerns. Generally, social desirability should be more of a concern for variables in which employees are being asked to say positive or negative things about themselves. In the present study, the common correlates measured via self-report differ in the degree to which they should be affected by socially desirable responding. For instance, among the Big Five personality variables, Impression Management

scale scores are most strongly related to Conscientiousness, Agreeableness, and Emotional Stability and less so to Extraversion and Openness (Berry, Page, & Sackett, 2007; Li & Bagger, 2006; Paulhus & Johns, 1998). Self-reports of job satisfaction (Ones & Viswesvaran, 1998) and negative affect (Chen, Dai, Spector, & Jex, 1997) have also been linked to socially desirable responding. Furthermore, because most employees likely desire to see themselves as good organizational citizens, self-reports of organizational citizenship behaviors are also suspect in terms of socially desirable responding. In contrast, social desirability should be less of a concern for variables in which respondents are asked to report objective information about themselves (i.e., demographics) or to provide ratings regarding aspects of their organizational context (i.e., organizational justice, conflict in the workplace, or organizational constraints employees encounter). Therefore, to the degree that self-report CWB is contaminated by socially desirable responding, self- and other-report CWB can be expected to have less similar patterns of relationships with common correlates affected by social desirability (i.e., Conscientiousness, Agreeableness, Emotional Stability, job satisfaction, negative affect, and organizational citizenship behavior), and more similar patterns of relationships with common correlates that are less affected by social desirability (i.e., Openness, Extraversion, demographics, organizational justice, organizational conflict, and organizational constraints; see Table 1 for definitions of each of these variables). Hypothesis 2b: Self- and other-report CWB will have less similar patterns of relationships with theoretically relevant correlates affected by socially desirable responding than with theoretically relevant correlates relatively unaffected by socially desirable responding.

Do Self- and Other-Raters Report Similar Amounts of CWB for Employees? This represents another line of evidence for the usefulness of other-ratings of CWB. As stated before, if self- and other-report CWB each capture unique variance (Hypothesis 1), it is still only worth going to the trouble to collect other-ratings if the unique variance captured by other-ratings reflects useful, incremental information. If self- and other-report CWB reflect different information, then it is further expected that each source would report different mean amounts of CWB for employees. As was discussed above, there are a number of reasons to expect self- and other-raters to provide dissimilar mean ratings, including rating source differences in the opportunity to observe CWB, other-raters’ over and/or underreporting of an employee’s CWB, as well as self-raters’ underreporting of their own CWB. These issues are likely to influence self- and other-ratings differently, so it is not clear whether one source will systematically report more or less CWB than the other. Therefore, we investigated the following research question: Research Question 1: Is there a difference in the mean amount of CWB reported by self- and other-raters? The same moderator analyses that were carried out for the correlation analyses were also carried out for the mean differences analyses. These mean difference moderator hypotheses were based on the same logic as the correlation moderator hypotheses. Specifically:

SELF- AND OTHER-REPORTS OF CWB

617

Table 1 Definitions for, and Measures of, Each Variable Meta-Analyzed Variable Counterproductive work behavior (CWB) CWB

Big Five Emotional Stability

Extraversion Openness to Experience Agreeableness Conscientiousness Organizational citizenship behavior (OCB) OCB

Organizational justice Distributive justice Interactional justice Procedural justice Affective variables Job satisfaction Negative affect

Contextual variables Conflict Constraints Demographics Age Gender Tenure

Definition

Measures included in the meta-analysis

Voluntary employee behaviors that are viewed by the organization as contrary to its legitimate interests, violate significant organizational norms, and threaten the well-being of the organization or its members. Also includes interpersonal- (e.g., violence, gossip) and organizational target (working slowly, damaging company property) dimensions.

Antisocial behavior (Robinson & O’Leary-Kelly, 1998); anticitizenship behavior (Ball et al., 1994); Counterproductive Work Behavior Checklist (Fox et al., 2001; Spector et al., 2006); workplace deviance (Aquino et al., 1999); Workplace Deviance Scale (Bennett & Robinson, 2000)

The personality trait reflecting the degree to which a person is secure, is calm, has low anxiety, and has low emotionality.

Big Five Inventory (John et al., 1991); International Personality Item Pool (Goldberg, 1999); Personal Characteristics Inventory (Barrick & Mount, 1999) Measured using same inventories as Emotional Stability

The personality trait reflecting the degree to which a person is sociable, assertive, talkative, ambitious, and energetic. The personality trait reflecting the degree to which a person is curious, intelligent, imaginative, and independent. The personality trait reflecting the degree to which a person is likable, easy to get along with, and friendly. The personality trait reflecting the degree to which a person is hard working, dependable, and detail oriented.

Measured using same inventories as Emotional Stability Measured using same inventories as Emotional Stability Measured using same inventories as Emotional Stability

Work behaviors that support the broader organizational, social, and psychological environment. Also includes interpersonal- and organizational target dimensions.

Lee & Allen (2002); Podsakoff et al. (1990); L. J. Williams & Anderson (1991)

The degree to which an employee feels the allocation of outcomes was fair. The degree to which an employee feels he or she has been treated sensitively and is respected. Degree to which an employee feels the process by which outcomes are distributed was fair.

Colquitt (2001); Price & Mueller (1986) Colquitt (2001); Niehoff & Moorman (1993) Colquitt (2001); Moorman (1991); Niehoff & Moorman (1993)

Pleasurable emotional state resulting from the appraisal of one’s job as achieving or facilitating the achievement of one’s job values. Negative emotions or subjectively experienced negative feelings.

Brayfield & Rothe (1951); Cammann et al. (1979)

Discrepant views or perceived incompatibilities between individuals. Situations/things at work that prevent employees from performing. Age (in years) of the participant 0 _ female, 1 _ male The number of years an employee has been employed by an organization.

Job-related Affective Well Being Scale (Van Katwyk et al., 2000); Positive and Negative Affect Schedule (Watson et al., 1988) Interpersonal Conflict at Work Scale (Spector & Jex, 1998) Organizational Constraints Scale (Spector & Jex, 1998)

BERRY, CARPENTER, AND BARRATT

618

Hypothesis 3a: There will be a larger mean difference in CWB ratings between supervisors and employees than between coworkers and employees. Hypothesis 3b: There will be a larger mean difference between self- and other-ratings of organizational target CWB than interpersonal target CWB. Hypothesis 3c: The mean difference between self- and otherratings of CWB will decrease as the assurances of participants’ anonymity increase.

Method Search for Primary Data A fivefold approach was used to identify studies containing useful information for the present meta-analysis. First, keyword searches of the PsycINFO and ABI/Inform databases were performed using 29 different keywords (see Table A1 in the Appendix for the search terms). Second, the Web of Science was used to identify any articles that cited Bennett and Robinson (2000). Third, manual searches of the 2005–2010 programs for Society for Industrial and Organizational Psychology and Academy of Management conferences were carried out. Fourth, the reference sections of relevant meta-analyses (Berry, Ones, & Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007; Lau, Au, & Ho, 2003; Salgado, 2002) were examined for useful citations. Fifth, prominent CWB researchers were contacted with requests for any unpublished research. Studies that contained enough information to extract either (a) a correlation between self-report CWB and other-report CWB (i.e., supervisor or coworker reports of CWB) or (b) a correlation between other-report CWB and some other variable were initially included. Only relationships based on at least three independent samples were meta-analyzed (this was the same decision rule used in Berry, Ones, & Sackett, 2007, and Cohen-Charash & Spector, 2001). This resulted in a final database of 40 studies from which 50 independent samples containing 224 independent correlations were drawn (see Table A2 in the Appendix for data regarding the primary studies). Twenty-one variables were identified that met the inclusion criteria (see Table 1 for a listing and definitions of each of these variables, along with the measures of each variable that were included in the meta-analysis). Of these 21 variables, 18 were correlates of CWB that will hereafter be collectively referred to as the CWB correlates.

Coding of Study Characteristics For each sample, the correlation between (a) self- and otherreports of CWB and/or (b) the correlation between other-reports of CWB and CWB correlates was coded. In cases in which relationships with the overall CWB construct were of interest but multiple facets of the overall construct were offered within the same sample (e.g., both interpersonal- and organizational target CWB measures, which are multiple facets of overall CWB, listed for the same sample), composite formulas (Ghiselli, Campbell, & Zedeck, 1981, pp. 163–164) were used to estimate correlations with a composite of the multiple measures. The mean and standard deviation of CWB reported by self- and other-raters was also coded for use in calculating standardized mean differences in CWB

between self- and other-ratings. It should be noted that an explicit search for studies reporting mean differences between self- and other-reported CWB was not carried out; instead, mean differences were coded in the correlational studies that provided relevant means and standard deviations. Two potential moderating variables were coded. First, the source of the other-rating (either supervisor or coworker) was coded as a potential moderator variable. Second, the procedures used by primary study researchers to assure participants of their anonymity were coded. In no studies were participant responses completely anonymous; at the very least, there must be identifiers to link self-reports of CWB to other-reports of CWB. However, three procedures were commonly cited in primary studies as methods for assuring participants of anonymity of responses: (a) Participants mailed completed paper-and-pencil surveys directly to the researchers (as opposed to other practices with paper-andpencil surveys such as having supervisors and employees simultaneously fill out and return the surveys during the workday), (b) self-raters generated a secret code that was provided to the otherrater, and/or (c) ratings were provided online.2 Some primary studies did not report using any of these “anonymity safeguards,” whereas some studies reported using as many as two. Thus, primary studies were sorted into three different categories representing the number of these three practices that were reported as having been used in the study (0, 1, or 2), based on the idea that using more of these practices would make participants feel their responses were more confidential. Efforts were also made to code for a number of other potential moderators (e.g., opportunity raters had to observe employees engaging in CWB, whether self- and other-raters were rating the same set of CWBs, whether Likert vs. frequency CWB scales were used, sample characteristics such as age and tenure), but primary studies either did not report relevant information often enough, or there was no variability across primary studies on the moderator variable. The second and third authors each independently coded five articles containing 82 correlations and 740 pieces of coding information and then met to quantify levels of intercoder agreement. The second and third authors agreed 100% on all variables. After this point, the second and third authors divided the rest of the articles for coding.

Analyses Hunter and Schmidt’s (2004) meta-analysis methods were used for meta-analyses of correlations and d values. Corrections for two statistical artifacts were made. First, point-biserial correlations involving the variable “gender” were individually corrected to what they would be if each sample had a 50 –50 gender split (see Table A2 for gender splits for samples). Second, correlations and 2

Of course, providing ratings online does not assure participants’ anonymity. More important is whether completing measures online enhances participants’ perceptions that their responses are anonymous, and there is evidence that this is the case. For instance, Richman, Kiesler, Weisband, and Drasgow’s (1999) meta-analysis demonstrated that social desirability distortion was reduced when measures were completed using a computer compared with a paper-and-pencil format as well as to face-to-face interviews. This suggests that completing surveys on a computer may lead respondents to feel more comfortable providing information that may be of a sensitive nature or socially undesirable.

SELF- AND OTHER-REPORTS OF CWB

d values were corrected for unreliability in both the predictor and the criterion3 using the artifact distribution method (see Table 2 for predictor and criterion reliability artifact distributions and their sources). For moderator analyses, studies were sorted into different categories on the basis of study characteristics (e.g., whether other-ratings were from supervisors vs. coworkers), and meta-analyses were carried out within each moderator category. To determine whether correlations or d values (which were converted to point-biserial correlations for the significance analyses) differed significantly across moderator categories, formulas from Raju and Brand (2003; specifically, Formulas 9 and 14) were used to test the significance of the difference between the corrected meta-analytic correlations. For the meta-analyses involving correlations between otherreports of CWB and (a) organizational citizenship behavior, (b) organizational constraints, and (c) interpersonal conflict, it was common for the other-raters to rate both variables involved in the correlation (e.g., supervisor raters rating both employees’ CWB and organizational citizenship behavior [OCB]). Thus, common method bias was a concern for such samples. Therefore, for the above relationships, analyses were run two ways: (a) including the samples in which other-raters rated both variables and (b) excluding such common-source samples. Meta-analytic correlations were much higher when such common-source samples were included. So, throughout the rest of the present article, the focus is directed toward analyses excluding such common-source samples; although, results with and without such samples are reported in the tables included in the present article.

Results Relationship Between Self- and Other-Ratings of CWB Hypothesis 1 suggested that self- and other-ratings of CWB would be positively correlated, but less than unity. Table 3 lists results for the relationship between self-ratings and other-ratings of CWB. The average corrected self-other CWB correlation was .38. Both the confidence and credibility intervals were well above zero and did not overlap with 1.0. This moderate positive correlation supports Hypothesis 1. For the following moderator analyses (Hypotheses 1a–1c), the number of samples in some categories is relatively small, so second-order sampling error may affect results. Hypothesis 1a suggested that self-ratings of CWB would be more strongly correlated with coworker ratings of CWB than supervisor ratings of CWB. Table 3 also lists the corrected correlations between selfand supervisor ratings of CWB (.37) and between self- and coworker ratings of CWB (.40). These correlations did not significantly differ (z ⫽ ⫺.82), and therefore, Hypothesis 1a was not supported. Hypothesis 1b suggested that self- and other-ratings of interpersonal target CWB (CWB-I) would correlate more strongly than self- and other-ratings of organizational target CWB (CWB-O). Table 3 lists meta-analytic correlations between self- and otherratings of CWB-I and CWB-O. The corrected correlation between self- and other-ratings of CWB-I was .51, whereas the corrected correlation between self- and other-ratings of CWB-O was .35. These correlations differed significantly (z ⫽ 4.35), providing support for Hypothesis 1b.4 Hypothesis 1c suggested that self- and

619

other-ratings of CWB would correlate more strongly as more steps were taken by researchers to assure participants of their anonymity. All but one sample included at least one coded “anonymity safeguard,” so analyses could only be carried out for samples including one versus two anonymity safeguards. The self– other CWB correlation was significantly greater in the two safeguards samples (.44) than in the one safeguard samples (.28; z ⫽ ⫺3.29). Thus, Hypothesis 1c was supported.

Comparison of the Relationships That Self-Report CWB and Other-Report CWB Have With Their Common CWB Correlates Table 4 lists full meta-analytic results for the correlations between other-ratings of CWB and the set of CWB correlates.5 Table 5 provides a side-by-side comparison of just the correlations 3 The criterion in the present meta-analysis was other-ratings of CWB, which involved judgmental ratings about employees’ CWB from supervisors or coworkers. There has been debate about whether interrater reliability versus alpha reliability should be used for correcting correlations in such instances (Schmidt, Viswesvaran, & Ones, 2000, argue that interrater reliability is most appropriate; other researchers have suggested that interrater reliabilities underestimate reliability; Murphy & DeShon, 2000). Therefore, the present meta-analysis offers separate estimates corrected for interrater reliability versus alpha reliability in the dependent variable for comparison purposes. One of the main purposes of the present metaanalysis was to compare the present study’s other-report CWB results with results from previous meta-analyses using self-reports (which were corrected using alpha coefficients). So, for the sake of comparability with the previous meta-analyses in which alpha coefficients were used in CWB corrections, the results in our tables are based on the correlations corrected using alpha coefficients, and we mostly focus our discussion on those estimates. However, within each table of results, there is a column listing the correlation corrected for interrater reliability in the dependent variable. Comparing the alpha- versus interrater-corrected estimates shows that the pattern of results is exactly the same, but the magnitude of correlations is slightly higher for the interrater-corrected estimates. 4 We note that the confidence intervals and credibility intervals for the self– other CWB-I and CWB-O correlations overlap (this is also the case with the other statistically significant moderator analyses). This suggests that, although the difference in the mean correlations is likely greater than zero, there is still some noteworthy overlap of the sampling error distributions for each individual corrected mean correlation (as captured by the confidence intervals) and the distributions of corrected mean correlations (as captured by the credibility intervals). 5 Attempts were also made to carry out moderator analyses testing whether the relationships that other-report CWB had with the common correlates differed depending on (a) whether other-ratings were from supervisors versus coworkers, (b) whether the other-ratings were of CWB-I versus CWB-O, or (c) number of anonymity safeguards. In most cases, there were not enough samples in each moderator category for meaningful moderator analyses (i.e., there were fewer than three samples). There were enough samples to carry out moderator analyses of the relationships other-report CWB had with job satisfaction, age, gender, and tenure; however, there was not enough variance in the results reported in Table 4 for meaningful moderator analyses for job satisfaction and age. So, moderator analyses were carried out only for gender and tenure, but neither moderating variable moderated the relationships between other-report CWB and gender and tenure. Therefore, there is no evidence that the relationships between other-report CWB and the common correlates are moderated by the variables included in the present study.

BERRY, CARPENTER, AND BARRATT

620

Table 2 Alpha and Interrater Reliability Artifact Distributions Variable CWB CWB (self-rating) CWB-I (self-rating) CWB-O (self-rating) CWB (other-rating) CWB-I (other-rating) CWB-O (other-rating) CWB (other-rating) CWB-I (other-rating) CWB-O (other-rating) Big Five Emotional Stability Extraversion Openness to Experience Agreeableness Conscientiousness OCB Overall OCB OCB-I OCB-O Organizational justice Distributive justice Interactional justice Procedural justice Affective variables Job satisfaction Negative affect Contextual variables Conflict Constraints

Type

rxx

SD

N

k

Source

Alpha Alpha Alpha Alpha Alpha Alpha Interrater Interrater Interrater

.77 .84 .82 .91 .88 .89 .665 .665 .665

— 0.07 0.07 0.05 0.07 0.06 — — —

16,721 6,878 6,080 3,411 4,594 4,958 1,125 1,125 1,125

49 26 22 21 26 28 13 13 13

Alpha Alpha Alpha Alpha Alpha

.78 .78 .73 .75 .78

0.11 0.09 0.12 0.11 0.10

— — — — —

370 307 251 123 307

Viswesvaran Viswesvaran Viswesvaran Viswesvaran Viswesvaran

Alpha Alpha Alpha

.79 .73 .74

— — —

16,455 5,864 5,607

47 24 23

Dalal (2005) Dalal (2005) Dalal (2005)

Alpha Alpha Alpha

.91 .90 .91

— 0.06 —

— 668 —

66 3 66

Hauenstein et al. (2001) Berry, Ones, & Sackett (2007) Hauenstrein et al. (2001)

Alpha Alpha

.84 .79

0.06 0.11

1,415 3,326

7 15

Connolly & Viswesvaran (2000) Connolly & Viswesvaran (2000)

Alpha Alpha

.74 .85

— —

3,363 1,746

13 8

Spector & Jex (1998) Spector & Jex (1998)

Dalal (2005) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Present study Present study Present study Viswesvaran et al. (1996) Viswesvaran et al. (1996) Viswesvaran et al. (1996) & & & & &

Ones Ones Ones Ones Ones

(2000) (2000) (2000) (2000) (2000)

Note. The mean and standard deviation of interrater reliabilities drawn from Viswesvaran et al. (1996) are averages of Viswesvaran et al.’s supervisor and peer interrater reliabilities. Type ⫽ type of reliability coefficient used in the artifact distribution; rxx ⫽ reliability artifact distribution mean; SD ⫽ reliability artifact distribution standard deviation; N ⫽ reliability artifact distribution sample size; k ⫽ number of samples contributing to artifact distributions; Source ⫽ study from which artifact distributions were drawn; CWB ⫽ counterproductive work behavior; CWB-I ⫽ interpersonal target CWB; CWB-O ⫽ organizational target CWB; OCB ⫽ organizational citizenship behavior; OCB-I ⫽ interpersonal target OCB; OCB-O ⫽ organizational target OCB. Dashes represent instances in which that piece of information was unavailable.

that self-report CWB and other-report CWB have with their common CWB correlates. The other-report-corrected correlations are each taken from Table 4 in the present meta-analysis; alphacorrected correlations appear outside and interrater-corrected correlations appear inside parentheses. The self-report-corrected correlations are taken from three previous meta-analyses (Berry, Ones, & Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007) in which the same relationships were examined, but relied almost exclusively on self-reports of CWB. There are two trends in Table 5 that are of importance: (a) the degree to which the magnitude of relationships differs depending on the source of CWB ratings (this deals most directly with Hypothesis 2a) and (b) the degree to which the pattern of relationships differs depending on the source of CWB ratings (most directly relevant to Hypothesis 2b). Regarding the magnitude of relationships, self-report CWB tended to be slightly more related to CWB correlates than otherreport CWB. The last column in Table 5 lists the difference between the self-report CWB and other-report CWB correlations (numbers outside parentheses used alpha-corrected other-report CWB for this calculation, numbers inside parentheses used interrater-corrected other-report CWB), with positive numbers meaning the self-report correlation is stronger. In 13 out of 18

relationships, the self-report correlation was stronger than the other-report correlation. In some instances, the correlation with self-report CWB was quite a bit stronger than the correlation with other-report CWB. For instance, organizational target OCB (OCB-O) was correlated ⫺.44 and ⫺.13 with self- and otherreport CWB, respectively; job satisfaction was correlated ⫺.37 and ⫺.21 with self- and other-report CWB, respectively. However, there were also instances in which the other-report correlation was quite a bit stronger than the self-report correlation (e.g., interpersonal target OCB [OCB-I], interactional justice). Furthermore, the overall magnitude of relationships did not differ especially strongly between self-report and other-report CWB, with the average difference across all common correlates being .07 (selfreport stronger) when using the alpha-corrected other-report correlations. When the interrater-corrected correlations are used, this difference reduces to only .04. Thus, there is, at best, weak and sporadic support for Hypothesis 2a that self-ratings of CWB will correlate more strongly than other-ratings of CWB with a common set of theoretically relevant correlates, suggesting common method bias is not a large systematic concern. Regarding the pattern of relationships, Hypothesis 2b suggested that self- and other-report CWB may show more similar relation-

SELF- AND OTHER-REPORTS OF CWB

621

Table 3 Meta-Analysis Results for the Relationship Between Self-Report and Other-Report Counterproductive Work Behavior Variable Overall analysis Self–other Moderator analyses Supervisor vs. coworker ratings Self-supervisor Self-coworker Interpersonal vs. organizational target CWB Self CWB-Other CWBI Self CWBO-Other CWBO Number of anonymity safeguards One Two

N

k

rm

SDr

r␣

SDr␣

r␣,irr

% var.

CV10

CV90

CIL

CIU

z

3,503

21

.32

.16

.38

.18

.44

18.51

.16

.61

.30

.47

2,044 1,459

11 10

.31 .33

.19 .12

.37 .40

.21 .11

.43 .47

12.61 40.74

.10 .26

.64 .54

.23 .31

.50 .49

1,500 1,500

9 9

.44 .29

.14 .17

.51 .35

.14 .18

.58 .40

20.63 18.74

.32 .12

.69 .58

.40 .22

.61 .48

4.35ⴱ

1,128 868

6 7

.23 .37

.08 .09

.28 .44

.06 .04

.32 .51

66.93 82.26

.20 .38

.35 .50

.20 .36

.36 .52

⫺3.29ⴱ

⫺0.82

Note. rm ⫽ mean sample size-weighted correlation; SDr ⫽ sample size-weighted observed standard deviation of correlations; r␣ ⫽ mean sample size-weighted correlation corrected for unreliability using alphas; SDr␣ ⫽ corrected standard deviation of corrected correlations; r␣,irr ⫽ mean sample size-weighted correlation corrected for unreliability using interrater reliabilities for other-rated variables and alphas for self-rated variables; % var. ⫽ percentage of variance attributable to statistical artifacts; CV10 and CV90 ⫽ 10% and 90% credibility values, respectively; CIL and CIU ⫽ lower and upper bounds, respectively, of the 95% confidence interval around the corrected mean correlation; z ⫽ the z statistic calculated using Formula 14 (and using Formula 9 to calculate sampling variance) from Raju and Brand (2003) for assessing the significance of the difference between the corrected correlations within each moderator category (zs ⱖ ⫾1.96 suggest a significant difference). Self CWBI-Other CWBI ⫽ correlation between self-ratings and other-ratings of CWB-Interpersonal; Self CWBO-Other CWBO ⫽ correlation between self-ratings and other-ratings of CWB-Organizational. ⴱ p ⬍ .05.

ships with common correlates that are less affected by social desirability (i.e., Extraversion, Openness, demographics, organizational justice, organizational conflict, and organizational constraints). The average difference in the last column in Table 5 (using alpha-corrected correlations for other-report CWB) for variables relatively unaffected by social desirability was .02, whereas it was .13 for the variables that were relatively affected by social desirability (the average differences were ⫺.01 and .10, respectively, when using interrater-corrected correlations). This provides some support for Hypothesis 2b. However, even differences in correlations of around .10 –.13 are not especially large (e.g., Berry, Ones, & Sackett, 2007). Furthermore, it is worth noting that the overall pattern of relationships that self-report CWB and otherreport CWB have with the CWB correlates is very similar. For instance, the vector of other-report CWB correlations was correlated .87 with the vector of self-report CWB correlations in Table 5 (these correlations were .93 and .77 for the variables relatively unaffected and affected by social desirability, respectively). These correlations are the same, regardless of whether alpha-corrected or interrater-corrected correlations are used. Thus, although self- and other-reports of CWB do have less similar patterns of relationships with their common correlates when the common correlates are relatively affected by social desirability, and the overall patterns of relationships are very similar for self- and other-report CWB.

Mean Differences Between Self- and Other-Ratings of CWB Mean difference results are presented in Table 6. Positive d values in Table 6 indicate self-raters reported more CWB. Research Question 1 asked whether there was a difference in the mean amount of CWB reported by self- and other-raters. The corrected mean difference was d ⫽ 0.35, and the confidence interval did not overlap with zero, suggesting that self-raters report more CWB than other-raters.

For the following moderator analyses (Hypotheses 1a–1c), the number of samples in some categories is relatively small, so second-order sampling error may affect results. Hypothesis 3a suggested there will be a larger mean difference in CWB ratings between supervisors and employees than between coworkers and employees. There was a significantly larger mean difference between supervisors and employees (d ⫽ 0.44) than between coworkers and employees (d ⫽ 0.23; z ⫽ 2.17), providing support for Hypothesis 3a. Hypothesis 3b suggested there will be a larger mean difference between self- and other-ratings of CWB-O than of CWB-I. There was a significantly larger mean difference between self- and otherratings of CWB-O (d ⫽ 0.46) than of CWB-I (d ⫽ 0.18; z ⫽ ⫺2.80), providing support for Hypothesis 3b. Hypothesis 3c suggested that the mean difference between selfand other-ratings of CWB will decrease as the assurances of participants’ anonymity increase. The self– other CWB mean difference was significantly larger in samples with one anonymity safeguard (d ⫽ 0.49) than in samples with two anonymity safeguards (d ⫽ 0.26; z ⫽ 2.08), providing support for Hypothesis 3c.

Is the Unique Variance Captured by Self- and Other-Report CWB Valid? The moderate correlation between self- and other-report CWB (.38) suggests that each source captures unique variance. What is not clear is whether the unique variance captured by self- and other-reports represents valid, unique variance (i.e., true CWB variance). One relevant line of evidence is the correlations that self- and other-report CWB have with the common correlates once variance attributable to the other source has been partialed out (e.g., what is the partial correlation between self-report CWB and a given common correlate when other-report variance has been partialed out of both self-report CWB and the common correlate?). If other-report

BERRY, CARPENTER, AND BARRATT

622

Table 4 Meta-Analysis Results for the Relationship Between Other-Report Counterproductive Work Behavior and the CWB Correlates Variable Big Five Emotional Stability Extraversion Openness to Experience Agreeableness Conscientiousness OCB Overall OCB Overall OCB* OCB-I OCB-I* OCB-O OCB-O* Organizational justice Distributive justice Interactional justice Procedural justice Affective variables Job satisfaction Negative affect Contextual variables Conflict Conflict* Constraints Constraints* Demographics Age Gender Tenure

N

k

rm

SDr

r␣

SDr␣

rirr,␣

% var.

CV10

CV90

CIL

CIU

2,975 1,066 890 2,246 3,332

12 7 6 9 13

⫺.04 .03 ⫺.10 ⫺.18 ⫺.15

.09 .13 .06 .16 .13

⫺.05 .04 ⫺.13 ⫺.22 ⫺.18

.07 .13 .00 .18 .14

⫺.06 .04 ⫺.15 ⫺.25 ⫺.21

53.09 36.35 100 14.94 22.03

⫺.14 ⫺.13 ⫺.13 ⫺.44 ⫺.36

.04 .20 ⫺.13 .01 ⫺.01

⫺.11 ⫺.08 ⫺.19 ⫺.34 ⫺.27

.01 .16 ⫺.07 ⫺.09 ⫺.10

1,509 712 2,016 612 1,065 612

11 5 9 4 7 4

⫺.47 ⫺.16 ⫺.33 ⫺.27 ⫺.46 ⫺.11

.17 .09 .18 .18 .13 .09

⫺.56 ⫺.19 ⫺.41 ⫺.34 ⫺.56 ⫺.13

.19 .03 .21 .20 .13 .06

⫺.65 ⫺.23 ⫺.48 ⫺.39 ⫺.66 ⫺.15

15.41 91.95 10.98 17.97 25.85 72.10

⫺.80 ⫺.23 ⫺.68 ⫺.59 ⫺.74 ⫺.21

⫺.32 ⫺.16 ⫺.14 ⫺.08 ⫺.39 ⫺.05

⫺.68 ⫺.28 ⫺.56 ⫺.55 ⫺.68 ⫺.24

⫺.43 ⫺.11 ⫺.27 ⫺.12 ⫺.45 ⫺.02

820 505 1,659

5 3 10

⫺.07 ⫺.40 ⫺.20

.20 .21 .18

⫺.07 ⫺.45 ⫺.22

.21 .22 .18

⫺.08 ⫺.52 ⫺.26

14.49 9.36 16.75

⫺.34 ⫺.73 ⫺.46

.19 ⫺.16 .01

⫺.27 ⫺.71 ⫺.35

.13 ⫺.18 ⫺.10

2,231 761

13 5

⫺.19 .16

.08 .08

⫺.21 .19

.04 .01

⫺.25 .23

82.63 99.28

⫺.26 .18

⫺.16 .20

⫺.26 .11

⫺.16 .28

1,609 934 1,262 838

12 7 9 6

.39 .31 .27 .20

.14 .09 .13 .09

.48 .38 .31 .22

.15 .05 .12 .04

.56 .44 .36 .26

26.99 75.86 36.79 85.58

.30 .31 .16 .17

.67 .44 .46 .27

.38 .29 .21 .14

.58 .46 .41 .30

2,091 2,080 1,552

13 13 10

⫺.05 ⫺.07 ⫺.06

.07 .15 .11

⫺.06 ⫺.08 ⫺.06

.00 .13 .07

⫺.07 ⫺.09 ⫺.07

100 27.21 55.63

⫺.06 ⫺.25 ⫺.16

⫺.06 .10 .03

⫺.10 ⫺.16 ⫺.13

⫺.02 .01 .01

Note. CWB ⫽ counterproductive work behavior; rm ⫽ mean sample size-weighted correlation; SDr ⫽ sample size-weighted observed standard deviation of correlations; r␣ ⫽ mean sample size-weighted correlation corrected for unreliability using alphas; SDr␣ ⫽ corrected standard deviation of corrected correlations; rirr,␣ ⫽ mean sample size-weighted correlation corrected for unreliability using interrater reliabilities for other-rated variables and alphas for self-rated variables; % var. ⫽ percentage of variance attributable to statistical artifacts; CV10 and CV90 ⫽ 10% and 90% credibility values, respectively; CIL and CIU ⫽ lower and upper bounds, respectively, of the 95% confidence interval around the corrected mean correlation. Asterisks denote meta-analyses wherein same-source samples were removed (e.g., supervisors rating both employees’ organizational citizenship behavior and CWB).

CWB no longer correlates with its common correlates once variance from self-reports has been partialed out, then this would suggest that other-reports tell us nothing about employees’ CWB that self-reports do not already tell us. The same would be true of self-report CWB’s partial correlations. Table 7 lists bivariate and partial correlations that self-report and other-report CWB have with the common correlates. Variance between self- and other-report correlations was partialed out using the meta-analytic correlation of .38 between self- and other-report CWB. Asterisks denote correlations with confidence intervals that do not overlap with zero. Fifteen of 18 partial correlations for self-report CWB and 10 of 18 partial correlations for other-report CWB had confidence intervals that did not overlap with zero, suggesting that both self- and other-reports contain at least some valid unique variance. However, many of the other-report CWB partial correlations, despite having confidence intervals that do not overlap with zero, are relatively small (around .10 or lower). In order to assess the value that is added by measuring CWB via both selfand other-ratings (i.e., multisource ratings) of CWB instead of just via self-ratings of CWB, each of the common correlates was hierarchically regressed on self-report CWB, and then on self-report and other-report CWB in a second step. The increase

in R2 from adding other-reports indicates the value of adding other-reports of CWB (see last column of Table 7). The increases in R2 were generally very small, only exceeding .02 in three instances (this pattern was the same if R, instead of R2 is used; increases in R ranged from .00 to .17, with only three increases being greater than .03).6 The three instances in which other-report CWB accounted for more than a .02 increase in R2 are noteworthy in that they involved the three most interpersonal CWB correlates: OCB-I, interactional justice, and conflict. In all, this suggests that the valid, unique variance added by other-report CWB over self-report CWB is generally very small, except perhaps when the CWB correlate is especially interpersonal in nature.

6

All of the analyses in Table 7 were also carried out (a) using observed, uncorrected correlations and (b) using interrater reliability-corrected correlations for any correlations involving other-report CWB. Results were virtually identical and no conclusions changed, so the observed correlation and interrater reliability-corrected correlation results are not listed. These results are available upon request from the first author.

SELF- AND OTHER-REPORTS OF CWB

623

Table 5 Side-by-Side Comparison of the Relationships That Other-Report CWB (Present Meta-Analysis) and Self-Report CWB (Previous Meta-Analyses) Have With a Common Set of CWB Correlates Variable Big Five Emotional Stability Extraversion Openness to Experience Agreeableness Conscientiousness OCB Overall OCB* OCB-I* OCB-O* Organizational justice Distributive justice Interactional justice Procedural justice Affective variables Job satisfaction Negative affect Contextual variables Conflict* Constraints* Demographics Age Gender Tenure

Correlation with other-report CWB

Correlation with self-report CWB

Source of self-report CWB correlation

Differencea

⫺.05 (⫺.06)b .04 (.04) ⫺.13 (⫺.15) ⫺.22 (⫺.25) ⫺.18 (⫺.21)

⫺.23 ⫺.03 ⫺.06 ⫺.35 ⫺.31

Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007)

.18 (.17) ⫺.07 (⫺.07) ⫺.07 (⫺.09) .13 (.10) .13 (.10)

⫺.19 (⫺.23) ⫺.34 (⫺.39) ⫺.13 (⫺.15)

⫺.36 ⫺.21 ⫺.44

Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007)

.17 (.13) ⫺.13 (⫺.18) .31 (.29)

⫺.07 (⫺.08) ⫺.45 (⫺.52) ⫺.22 (⫺.26)

⫺.14 ⫺.30 ⫺.27

Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007)

.07 (.06) ⫺.15 (⫺.22) .05 (.01)

⫺.21 (⫺.25) .19 (.23)

⫺.37 .29

Dalal (2005) Hershcovis et al. (2007)

.16 (.12) .10 (.06)

.46 .33

Hershcovis et al. (2007) Hershcovis et al. (2007)

.08 (.02) .11 (.07)

.38 (.44) .22 (.26) ⫺.06 (⫺.07) ⫺.08 (⫺.09) ⫺.06 (⫺.07)

⫺.13 ⫺.17 ⫺.05

Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007) Berry, Ones, & Sackett (2007)

.07 (.06) .09 (.08) ⫺.01 (⫺.02)

Note. CWB ⫽ counterproductive work behavior; OCB ⫽ organizational citizenship behavior; OCB-I ⫽ interpersonal target OCB; OCB-O ⫽ organizational target OCB. Asterisks denote common-source samples (e.g., supervisor rated both OCB and CWB) removed. a Positive numbers mean the correlation with self-report CWB was stronger than the correlation with other-report CWB. b Numbers outside parentheses reflect correlations corrected using alpha coefficients for other-report CWB; numbers inside parentheses reflect correlations corrected using interrater reliabilities for other-report CWB.

Discussion Summary of Findings The results of the present meta-analysis provided three lines of evidence regarding the comparability of self- and other-reports of CWB. First, self- and other-reports of CWB are moderately correlated (corrected correlations of .38 or .44, depending on whether alpha vs. interrater reliability is used for other-ratings). The correlation between self- and other-ratings becomes somewhat stronger (i.e., in the .40 –.60 range) when the ratings are of CWB-I and when greater assurances of anonymity are provided. This means that self- and other-ratings of CWB intercorrelate more strongly than ratings of job performance from multiple sources (Conway & Huffcutt, 1997). Thus, the general trend is for appreciable overlap between self- and other-ratings of CWB. However, the rating sources clearly each capture a fair amount of unique variance. This unique variance begs the question of whether self- and other-ratings of CWB have different patterns or magnitudes of relationships with their correlates. The second line of evidence presented in this meta-analysis demonstrated that, with some exceptions, self- and other-ratings of CWB have very similar relationships with their common correlates. The patterns of relationships that self- and other-report CWB have with their common correlates are almost identical, and self-reports of CWB correlated with their common correlates only slightly higher (between about

.04 and .07), on average, than other-reports. There are some noteworthy exceptions, such as the stronger relationships that self-report CWB has with Conscientiousness, Agreeableness, and Emotional Stability; the stronger relationships that other-report CWB has with the interpersonal correlates of OCB-I, interactional justice, and conflict; and the pattern of greater differences in relationships that self- and other-report CWB have with common correlates more affected by social desirability. However, even when both self-report CWB and the CWB correlates were likely to be affected by social desirability, self-report CWB only correlated with the common correlates about .10 –.13 higher than did otherreport CWB, which is a relatively small difference (e.g., Berry, Ones, & Sackett, 2007; Cohen, 1992). Additionally, the correlation of .87 between the vectors of self- and other-report correlations with the common correlates demonstrates that the general pattern is for the nomological networks of the two sources to be very similar. Furthermore, the partial correlation analyses demonstrated that for most of the CWB correlates, other-reports of CWB account for very little additional variance beyond self-reports of CWB. The third line of evidence regarding the comparability of selfand other-reports of CWB presented in this meta-analysis was the average mean difference between the amounts of CWB reported by the two sources. Across all conditions, self-raters reported engaging in more CWB than other-raters reported them engaging in. Thus, the idea that self-raters will underreport their engagement

BERRY, CARPENTER, AND BARRATT

624

Table 6 Mean Differences Between Self- and Other-Ratings of Employees’ CWB Variable Overall analysis Self–oOther Moderator analyses Supervisor vs. coworker ratings Self-supervisor Self-coworker Interpersonal vs. organizational target CWB Self CWBI-Other CWBI Self CWBO-Other CWBO Number of anonymity safeguards One Two

N

k

dm

SDd

␦␣

SD␦

␦␣,irr

% var.

2,574

17

.29

.25

.35

.22

.41

1,458 1,116

9 8

.37 .20

.27 .18

.44 .23

.26 .07

1,145 1,145

7 7

.15 .40

.24 .45

.18 .46

1,128 880

6 7

.41 .22

.25 .25

.49 .26

CV10

CV90

CIL

CIU

z

43.3

.07

.64

.21

.49

.52 .27

35.5 88.7

.12 .14

.77 .33

.23 .08

.65 .38

2.17ⴱ

.22 .49

.24 .54

41.1 12.6

⫺.10 ⫺.16

.46 1.09

⫺.03 .08

.38 .83

⫺2.80ⴱ

.24 .21

.58 .30

35.2 50.7

.19 ⫺.01

.80 .53

.26 .04

.73 .48

2.08ⴱ

Note. Positive d values indicate self-report means were higher; dm ⫽ mean sample size-weighted d value; SDd ⫽ sample size-weighted observed standard deviation of d values; ␦␣ ⫽ mean sample size-weighted d value corrected for measurement error using alpha reliability; SD␦ ⫽ corrected standard deviation of alpha-corrected d values; ␦␣,irr ⫽ mean sample size-weighted d value corrected for measurement error using alpha reliability for self-ratings and interrater reliability for other-ratings; % var. ⫽ percentage of variance attributable to artifacts; CV10 and CV90 ⫽ 10% and 90% credibility values, respectively; CIL and CIU ⫽ lower and upper bounds, respectively, of the 95% confidence interval around the corrected mean d value; z ⫽ the z statistic calculated using Formula 14 (and using Formula 9 to calculate sampling variance) from Raju and Brand (2003) for assessing the significance of the difference between the corrected correlations (d values were converted to point-biserial correlations) within each moderator category (zs ⱖ ⫾1.96 suggest a significant difference). Self CWBI-Other CWBI ⫽ mean difference between self-ratings and other-ratings of CWB-Interpersonal; Self CWBO-Other CWBO ⫽ mean difference between self-ratings and other-ratings of CWB-Organizational. ⴱ p ⬍ .05.

in CWB was not supported in the present meta-analysis—at least not in comparison to other-raters. However, the common concern that other-raters do not have adequate opportunity to observe employees engaging in CWB (and especially CWB-O), and therefore underreport employees’ CWB, was supported by the mean difference results of the present meta-analysis.

A number of hypotheses regarding moderating variables were also tested. With the exception of Hypothesis 1a, each of the hypotheses was supported. The identified moderating variables do not change general conclusions of the meta-analysis (e.g., although there was a larger mean difference between self- and supervisor ratings than between self- and coworker ratings, this does not

Table 7 Bivariate and Partial Correlations Between Self- and Other-Report CWB and the CWB Correlates CWB correlate

Bivariate-other

Partial-other

Bivariate-self

Partial-self

⌬R2

Emotional Stability Extraversion Openness to Experience Agreeableness Conscientiousness Overall OCB OCB-I OCB-O Distributive justice Interactional justice Procedural justice Job satisfaction Negative affect Conflict Constraints Age Gender Tenure

⫺.05 .04 ⫺.13* ⫺.22* ⫺.18* ⫺.19* ⫺.34* ⫺.13* ⫺.07 ⫺.45* ⫺.22* ⫺.21* .19* .38* .22* ⫺.06* ⫺.08 ⫺.06

.04 .06 ⫺.13* ⫺.10* ⫺.07* ⫺.06 ⫺.30* .04 ⫺.02 ⫺.39* ⫺.14* ⫺.08* .09* .24* .11* ⫺.01 ⫺.02 ⫺.05

⫺.23 ⫺.03 ⫺.06 ⫺.35* ⫺.31* ⫺.36* ⫺.21* ⫺.44* ⫺.14 ⫺.30* ⫺.27* ⫺.37* .29* .46* .33* ⫺.13* ⫺.17 ⫺.05

⫺.25* ⫺.05 ⫺.01 ⫺.31* ⫺.28* ⫺.34* ⫺.09* ⫺.46* ⫺.13* ⫺.15* ⫺.22* ⫺.34* .26* .37* .29* ⫺.13* ⫺.16* ⫺.03

.00* .00 .01* .01* .01* .00 .08* .00 .00 .13* .02* .01* .01* .05* .01* .00 .00 .00

Note. CWB ⫽ counterproductive work behavior; Bivariate-other ⫽ bivariate correlation between other-report CWB and the CWB correlate; Partialother ⫽ correlation between other-report CWB and a given CWB correlate, with self-report CWB variance partialed out of both variables; Bivariate-self ⫽ bivariate correlation between self-report CWB and the CWB correlate; Partial-self ⫽ correlation between self-report CWB and a given CWB correlate, with other-report CWB variance partialed out of both variables; ⌬R2 ⫽ the increase in the squared multiple correlation when regressing each common correlate on self- and other-ratings of CWB instead of just self-ratings of CWB. Asterisks mean that the confidence interval for the given correlation did not overlap with zero (using .05 alpha level). The meta-analytic sample sizes on which the correlations were based were used for determining whether the confidence interval overlapped with zero.

SELF- AND OTHER-REPORTS OF CWB

change the conclusion that other-raters report less CWB than self-raters). However, they do highlight some important points for future CWB research. First, supervisor raters report even less CWB, relative to self-raters, than do coworker raters, suggesting supervisors may have even less opportunity to observe employees engaging in CWB than coworkers. Second, self- and other-raters agree more about CWB-I than about CWB-O (i.e., ratings are more intercorrelated, and mean differences are smaller). Third, self- and other-ratings of CWB agree more as the number of anonymity safeguards increase, suggesting CWB researchers may get better data when participants perceive their responses as more anonymous.

Is it Worth the Trouble of Collecting Other-Ratings or Multisource Ratings of CWB? It is difficult to collect other-ratings or multisource ratings of CWB in organizations, especially as compared with collecting self-ratings of CWB. Thus, one might ask whether it is worth the trouble of collecting other-ratings or multisource ratings. The results of the present meta-analysis suggest in many instances that there is little tangible value of other-ratings of CWB (and by extension, multisource ratings) beyond selfratings. Other-ratings and self-ratings have mostly similar patterns and magnitudes of relationships with other variables, and other-ratings often do not account for appreciable incremental variance over self-ratings. Furthermore, the mean differences results of the present meta-analysis suggest that coworker, and especially supervisor, raters underreport employees’ CWB relative to the employees themselves, which suggests that the use of other-report CWB entails a focus on a narrower subset of employees’ CWB. Especially if researchers take multiple careful steps to assure participants of the anonymity of their responses, the value added by collecting other-reports of CWB is not clear in many instances. This is not to say that other-reports of CWB do not have value. For instance, the interpersonal CWB correlates OCB-I, interactional justice, and conflict were more strongly related to otherreport CWB than self-report CWB. This may suggest that supervisor and coworker raters have an especially valid perspective on CWBs that are interpersonal in nature or that have interpersonal antecedents (e.g., breaches of interactional justice, strong interpersonal conflict in the workplace). This conclusion must be tempered by concerns over second-order sampling error, as the meta-analytic relationships between other-report CWB and OCB-I, interactional justice, and conflict were based on four, three, and seven samples, respectively. However, the stronger correlation between self- and other-reports of the relatively interpersonal CWB-Is also corroborates the idea that other-reports may be most useful in assessing relatively public, interpersonal CWBs. Furthermore, the fact that patterns of results are so similar between self- and other-reports suggests that for many purposes, one source is as good as the other. So, if other-reports are available, then their use should generally be appropriate. Additionally, in some instances, such as test validation, organizations may view self-reports of CWB (or other criteria) skeptically. In such instances, other-reports may be a more viable option. However, the results of the present meta-analysis do not generally support the idea that self-reports are always an inferior method of assessing CWB.

625

Implications for Theory and Practice A practical implication of the results is that self- and otherratings of CWB generally result in very similar patterns of findings. Therefore, organizational researchers should feel justified in using either self- or other-ratings of CWB. Additionally, the findings of the present study indicate that supervisor and coworker ratings of CWB are relatively comparable (i.e., whether ratings came from supervisors vs. coworkers only moderated self– other mean differences, and the mean difference effect was relatively small [d ⫽ 0.21], although statistically significant). This agrees with the results of Viswesvaran, Schmidt, and Ones (2002), who found that supervisor and coworker ratings of “compliance/ acceptance of authority” (essentially avoidance of CWB) correlated .78. This is information future CWB researchers can draw on when deciding which rating sources are most appropriate, given their specific research question. Another practical implication flows from the largely similar patterns of relationships with correlates in nomological networks that self- and other-ratings of CWB exhibited. The scientific knowledge base to date regarding the causes and consequences of CWB that practitioners can draw on has largely depended on self-report measurement of CWB. The present meta-analysis suggests that use of other-report CWB does not drastically change conclusions regarding the causes and consequences of CWB. Thus, practitioners do not presently have cause for concern over the relevance or legitimacy of the self-report research-based CWB interventions they recommend to organizations. This also highlights perhaps the most important theoretical implication of the results of the present meta-analysis. Much of the recent empirical evidence supporting theories of CWB has relied on self-report measurement of CWB. In particular, much of what we know about CWB’s nomological network (e.g., Berry, Ones, & Sackett, 2007; Dalal, 2005; Hershcovis et al., 2007) has been based on self-report CWB measures. The present meta-analysis found that conclusions regarding the nomological network of CWB remain largely the same when otherreports of CWB are used. This validates much of the cumulative CWB literature and theory to date that has relied on multi-item self-report CWB measurement.

Additional Issues, Limitations, and Directions for Future Research Other lines of evidence regarding the convergence of selfand other-report CWB that were not available in the present meta-analysis are relevant and present opportunities for future research. For instance, an important line of evidence regarding the similarity/dissimilarity of self- and other-report CWB is the degree to which they exhibit similar factor structures (e.g., Dilchert, Ones, Davis, & Rostow, 2007). Thus, measurement invariance research would be useful. Additionally, the present meta-analysis was not able to directly test the degree to which self- and other-ratings of CWB are accurate (i.e., reflect “true CWB”) or the degree to which the shared variance between selfand other-ratings of CWB reflects true CWB variance versus shared, systematic error variance. Addressing this question would likely require experimental research, although the eco-

BERRY, CARPENTER, AND BARRATT

626

logical validity of such experimental CWB research might be difficult to establish. Also, although the amount of unique variance captured by other-ratings was relatively small in most instances, it may be of interest to determine the cause of this unique variance (e.g., what forms/types of CWB are otherraters more privy to observe their employees engaging in?). It is worth noting that the methods used in the present study are applicable not only to addressing issues regarding the comparability and nomological networks of self- and other-report CWB but also to any constructs in the broader research literature that face issues regarding the appropriateness of self- versus other-report measurement. For instance, the appropriateness of self- versus other-report measurement is relevant for many other constructs ranging from other performance behaviors (e.g., task performance, OCB, job/work withdrawal) to individual-differences measures (e.g., personality, adaptability) to contextual variables (e.g., organizational conflict, organizational constraints, organizational support) and beyond. We used a number of methods in the present meta-analysis to determine the relationship between self- and other-reports, the degree to which they have similar nomological networks, and the degree to which they account for unique versus overlapping variance. These same methods can, and perhaps should, be used for investigating similar issues with other variables such as those listed above.

References References marked with an asterisk indicate studies included in the meta-analysis. Aquino, K., Lewis, M. U., & Bradfield, M. (1999). Justice constructs, negative affectivity, and employee deviance: A proposed model and empirical test. Journal of Organizational Behavior, 20, 1073–1091. Ball, G. A., Trevino, L. K., & Sims, H. P. (1994). Just and unjust punishment: Influences on subordinate performance and citizenship. Academy of Management Journal, 37, 299 –322. Barclay, L. J., & Aquino, K. (2010). Workplace aggression and violence. In S. Zedeck (Ed.), APA handbook of industrial and organizational psychology (Vol. 3, pp. 615– 640). Washington, DC: American Psychological Association. Barrick, M. R., & Mount, M. K. (1999). The Personal Characteristics Inventory. Libertyville, IL: Wonderlic. Bennett, R. J., & Robinson, S. L. (2000). Development of a measure of workplace deviance. Journal of Applied Psychology, 85, 349 –360. doi: 10.1037/0021-9010.85.3.349 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, 410 – 424. doi: 10.1037/0021-9010.92.2.410 Berry, C. M., Page, R. C., & Sackett, P. R. (2007). Effects of self-deceptive enhancement on personality-job performance relationships. International Journal of Selection and Assessment, 15, 94 –109. doi:10.1111/ j.1468-2389.2007.00374.x *Bing, M., Stewart, S. M., Davison, H. K., Green, P. D., McIntyre, M. D., & James, L. R. (2007). An integrative typology of personality assessment for aggression: Implications for predicting counterproductive workplace behavior. Journal of Applied Psychology, 92, 722–744. doi: 10.1037/0021-9010.92.3.722 *Bordia, P., Restubog, S. L. D., & Tang, R. L. (2008). When employees strike back: Investigating mediating mechanisms between psychological contract breach and workplace deviance. Journal of Applied Psychology, 93, 1104 –1117. doi:10.1037/0021-9010.93.5.1104

Brayfield, A. H., & Rothe, H. F. (1951). An index of job satisfaction. Journal of Applied Psychology, 35, 307–311. *Bruk-Lee, V., & Spector, P. E. (2006). The social stressorscounterproductive work behaviors link: Are conflicts with supervisors and coworkers the same? Journal of Occupational Health Psychology, 11, 145–156. doi:10.1037/1076-8998.11.2.145 Budd, J. W., Arvey, R. D., & Lawless, P. (1996). Correlates and consequences of workplace violence. Journal of Occupational Health Psychology, 1, 197–210. doi:10.1037/1076-8998.1.2.197 Camara, W. J., & Schneider, D. L. (1994). Integrity tests: Facts and unresolved issues. American Psychologist, 49, 112–119. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. (1979). The Michigan Organizational Assessment Questionnaire. Unpublished manuscript, University of Michigan, Ann Arbor. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validity by the multitrait-multimethod matrix. Psychological Bulletin, 56, 81–105. doi:10.1037/h0046016 Chen, P. Y., Dai, T., Spector, P. E., & Jex, S. M. (1997). Relation between negative affectivity and positive affectivity: Effects of judged desirability of scale items and respondents’ social desirability. Journal of Personality Assessment, 69, 183–198. doi:10.1207/ s15327752jpa6901_10 *Chiu, S., & Peng, J. (2008). The relationship between psychological contract breach and employee deviance: The moderating role of hostile attributional style. Journal of Vocational Behavior, 73, 426 – 433. doi: 10.1016/j.jvb.2008.08.006 Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. doi:10.1037/0033-2909.112.1.155 Cohen-Charash, Y., & Spector, P. E. (2001). The role of justice in organizations: A meta-analysis. Organizational Behavior and Human Decision Processes, 86, 278 –321. doi:10.1006/obhd.2001.2958 *Colbert, A. E., Mount, M. K., Harter, J. K., Witt, L. A., & Barrick, M. R. (2004). Interactive effects of personality and perceptions of the work situation on workplace deviance. Journal of Applied Psychology, 89, 599 – 609. doi:10.1037/0021-9010.89.4.599 *Colbert, A. E., Witt, L. A., & Mount, M. K. (2003, April). Interactive effects of organizational support and agreeableness on interpersonal deviance. Poster presented at the meeting of Society for Industrial/ Organizational Psychology, Orlando, FL. Colquitt, J. A. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied Psychology, 86, 386 – 400. Connolly, J. J., & Viswesvaran, C. (2000). The role of affectivity in job satisfaction: A meta-analysis. Personality and Individual Differences, 29, 265–281. Conway, J. M., & Huffcutt, A. I. (1997). Psychometric properties of multisource performance ratings: A meta-analysis of subordinate, supervisor, peer, and self-ratings. Human Performance, 10, 331–360. doi: 10.1207/s15327043hup1004_2 Dalal, R. S. (2005). A meta-analysis of the relationship between organizational citizenship behavior and counterproductive work behavior. Journal of Applied Psychology, 90, 1241–1255. doi:10.1037/00219010.90.6.1241 *de Jonge, J., & Peeters, M. C. W. (2009). Convergence of self-reports and coworker reports of counterproductive work behavior: A crosssectional multi-source survey among health care workers. International Journal of Nursing Studies, 46, 699 –707. doi:10.1016/ j.ijnurstu.2008.12.010 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, 616 – 627. doi:10.1037/00219010.92.3.616 *Fallon, J. D., Avis, J. M., Kudisch, J. D., Gornet, T. P., & Frost, A. (2000). Conscientiousness as a predictor of productive and counterproductive

SELF- AND OTHER-REPORTS OF CWB behaviors. Journal of Business Psychology, 15, 339 –349. doi:10.1023/ A:1007880203956 *Ferris, D. L. (2007). [Correlations between self- and coworker-reports of counterproductive work behavior]. Unpublished raw data. *Ferris, D. L. (2008). [Correlations between self- and coworker-reports of counterproductive work behavior]. Unpublished raw data. *Ferris, D. L., Brown, D. J., Berry, J. W., & Lian, H. (2008). The development and validation of the Workplace Ostracism Scale. Journal of Applied Psychology, 93, 1348 –1366. doi:10.1037/a0012743 Fox, S., & Spector, P. E. (1999). A model of work frustration-aggression. Journal of Organizational Behavior, 20, 915–931. doi:10.1002/ (SICI)1099-1379(199911)20:6⬍915::AID-JOB918⬎3.0.CO;2-6 *Fox, S., Spector, P. E., Goh, A., & Bruursema, K. (2007). Does your coworker know what you’re doing? Convergence of self- and peerreports of counterproductive work behavior. International Journal of Stress Management, 14, 41– 60. doi:10.1037/1072-5245.14.1.41 Fox, S., Spector, P. E., & Miles, D. (2001). Counterproductive work behavior (CWB) in response to job stressors and organizational justice: Some mediator and moderator tests for autonomy and emotions. Journal of Vocational Behavior, 59, 291–309. doi:10.1006/ jvbe.2001.1803 Ghiselli, E. E., Campbell, J. P., & Zedeck, S. (1981). Measurement theory for the behavioral sciences. San Francisco, CA: W. H. Freeman & Co. Glomb, T. M. (2002). Workplace anger and aggression: Informing conceptual models with data from specific encounters. Journal of Occupational Health Psychology, 7, 20 –36. doi:10.1037/10768998.7.1.20 *Goh, A. (2007). An attributional analysis of counterproductive work behavior (CWB) in response to occupational stress (Unpublished doctoral dissertation). University of South Florida. Goldberg, L. R. (1999). A broad-bandwidth, public-domain, personality inventory measuring the lower-level facets of several Five-Factor models. In I. Mervielde, I. J. Deary, D. de Fruyt, & F. Ostendorf (Eds.), Personality psychology in Europe (pp. 7–28). Tilburg, the Netherlands. *Greenbaum, R. L., Bardes, M., Mayer, D., & Priesemuth, M. (2009, August). Supervisors’ abuse of customers: The moderating role of moral obligation. Paper presented at the Annual Meeting of Academy of Management, Chicago, IL. Hauenstein, N. M. A., McGonigle, T., & Flinder, S. W. (2001). A metaanalysis of the relationship between procedural justice and distributive justice: Implications for justice research. Employee Responsibilities and Rights Journal, 13, 39 –56. Heneman, H. G., Heneman, R. L., & Judge, T. A. (1997). Staffing organizations. Madison, WI: Mendota House/Irwin. Hershcovis, M. S., Turner, N., Barling, J., Arnold, K. A., Dupre, K. E., Inness, M., . . . Sivanathan, N. (2007). Predicting workplace aggression: A meta-analysis. Journal of Applied Psychology, 92, 228 –238. doi: 10.1037/0021-9010.92.1.228 *Hung, T. K., Chi, N. W., & Lu, W. L. (2009). Exploring the relationships between perceived coworker loafing and counterproductive work behaviors: The mediating role of a revenge motive. Journal of Business and Psychology, 24, 257–270. doi:10.1007/s10869-009-9104-6 Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis: Correcting error and bias in research findings (2nd ed.). New York, NY: Sage Publications. John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big Five Inventory: Versions 4a and 54 [Technical report]. Berkeley: Institute of Personality and Social Research, University of California, Berkeley. *Jones, D. A. (2004). Toward a better understanding of fairness in the workplace: Attitude strength, predictive asymmetry, and the revenge

627

motive (Unpublished doctoral dissertation). University of Calgary, Alberta, Canada. *Judge, T. A., LePine, J. A., & Rich, B. L. (2006a). Loving yourself abundantly: Relationship of the narcissistic personality to self- and other perceptions of workplace deviance, leadership, and task and contextual performance. Journal of Applied Psychology, 91, 762–776. doi:10.1037/ 0021-9010.91.4.762 *Judge, T., Scott, B. A., & Ilies, R. (2006b). Hostility, job attitudes, and workplace deviance: Test of a multilevel model. Journal of Applied Psychology, 91, 126 –138. doi:10.1037/0021-9010.91.1.126 *Kessler, S. R., Singla, N., Galperin, B. L., & Spector, P. E. (2009, April). Can there be too much of a good thing? Self-esteem and its impact on counterproductive work behavior. Poster presented at the meeting of Society for Industrial/Organizational Psychology, New Orleans, LA. *Kickul, J. R., Neuman, G., Parker, C., & Finkl, J. (2001). Settling the score: The role of organizational justice in the relationship between psychological contract breach and anticitizenship behavior. Employee Responsibilities and Rights Journal, 13, 77–93. doi:10.1023/A: 1014586225406 *Kidwell, R. E., & Bennett, N. (2001). Perceived work context and employee job neglect. American Business Review, 19, 64 –74. *Kuenzi, M., & Carson, J. (2009, August). Ethical leadership and power: Combined effects on unit outcomes. Poster presented at the meeting of the Academy of Management, Chicago, IL. Lau, V. C. S., Au, W. T., & Ho, J. M. C. (2003). A qualitative and quantitative review of antecedents of counterproductive behavior in organizations. Journal of Business and Psychology, 18, 73–99. *Lee, K. (2000). Job affect as a predictor of organizational citizenship behavior and workplace deviance (Unpublished doctoral dissertation). University of Western Ontario, Canada. *Lee, K., & Allen, N. J. (2002). Organizational citizenship behavior and workplace deviance: The role of affect and cognitions. Journal of Applied Psychology, 87, 131–142. doi:10.1037/0021-9010.87.1.131 Lee, R. M. (1993). Doing research on sensitive topics. London, England: Sage. Li, A., & Bagger, J. (2006). Using the BIDR to distinguish the effects of impression management and self-deception on the criterion validity of personality measures: A meta-analysis. International Journal of Selection and Assessment, 14, 131–141. doi:10.1111/j.1468-2389.2006 .00339.x Marcus, B., Schuler, H., Quell, P., & Humpfner, G. (2002). Measuring counterproductivity: Development and initial validation of a German self-report questionnaire. International Journal of Selection and Assessment, 10, 18 –35. doi:10.1111/1468-2389.00191 Moorman, R. H. (1991). Relationship between organizational justice and organizational citizenship behaviors: Do fairness perceptions influence employee citizenship? Journal of Applied Psychology, 76, 845– 855. Mount, M. K., Barrick, M. R., & Strauss, J. P. (1994). Validity of observer ratings of the Big Five personality dimensions. Journal of Applied Psychology, 79, 272–280. doi:10.1037/0021-9010.79.2.272 *Mount, M., Ilies, R., & Johnson, E. (2006). Relationship of personality traits and counterproductive work behaviors: The mediating effects of job satisfaction. Personnel Psychology, 59, 591– 622. doi:10.1111/ j.1744-6570.2006.00048.x Murphy, K. R. (1993). Honesty in the workplace. Pacific Grove, CA: Brooks/Cole. Murphy, K. R., & DeShon, R. (2000). Interrater correlations do not estimate the reliability of job performance ratings. Personnel Psychology, 53, 873–900. doi:10.1111/j.1744-6570.2000.tb02421.x Niehoff, B. P., & Moorman, R. H. (1993). Justice as a mediator of the relationship between methods of monitoring and organizational citizenship behavior. Academy of Management Journal, 36, 527–556.

628

BERRY, CARPENTER, AND BARRATT

*O’Brien, K. (2009). [Correlation between self- and peer-report counterproductive work behavior]. Unpublished raw data. *O’Brien, K. E., & Allen, T. D. (2007). The relative importance of correlates of organizational citizenship behavior and counterproductive work behavior using multiple sources of data. Human Performance, 21, 62– 88. doi:10.1080/08959280701522189 *O’Brien, K. E., & Loch, S. E. (2005, April). Development and correlates of the Workplace Hostile Attribution Bias Survey. Paper presented at the annual meeting of the Society for Industrial/Organizational Psychology, Los Angeles, CA. Ones, D. S., & Viswesvaran, C. (1998). The effects of social desirability and faking on personality and integrity assessment for personnel selection. Human Performance, 11, 245–269. Ones, D. S., Viswesvaran, C., & Schmidt, F. L. (1993). Comprehensive meta-analysis of integrity test validities: Findings from implications for personnel selection and theories of job performance. Journal of Applied Psychology, 78, 679 –703. Paulhus, D. L., & John, O. P. (1998). Egoistic and moralistic biases in self-perception: The interplay of self-deceptive styles with basic traits and motives. Journal of Personality, 66, 1025–1060. doi:10.1111/14676494.00041 *Penney, L. M., & Spector, P. E. (2005). Job stress, incivility, and counterproductive work behavior (CWB): The moderating role of negative affectivity. Journal of Organizational Behavior, 26, 777–796. doi: 10.1002/job.336 Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879 –903. doi:10.1037/0021-9010.88.5.879 Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers’ trust in leader, satisfaction, and organizational citizenship behaviors. Leadership Quarterly, 1, 107–142. Price, J. L., & Mueller, C. W. (1986). Absenteeism and turnover among hospital employees. Greenwich, CT: JAI Press. *Puffer, S. M. (1987). Prosocial behavior, noncompliant behavior, and work performance among commission salespeople. Journal of Applied Psychology, 72, 615– 621. doi:10.1037/0021-9010.72.4.615 Raju, N. S., & Brand, P. A. (2003). Determining the significance of correlations corrected for unreliability and range restriction. Applied Psychological Measurement, 27, 52–71. doi:10.1177/0146621602239476 *Richards, D., & Schat, A. (2011). [Correlation between self- and supervisor-ratings of counterproductive work behavior]. Unpublished raw data. Richman, W. L., Kiesler, S., Weisband, S., & Drasgow, F. (1999). A meta-analytic study of social desirability distortion in computeradministered questionnaires, traditional questionnaires, and interviews. Journal of Applied Psychology, 84, 754 –775. doi:10.1037/00219010.84.5.754 Robinson, S. L., & O’Leary-Kelly, A. M. (1998). Monkey see, monkey do: The influence of work groups on the antisocial behavior of employees. Academy of Management Journal, 41, 658 – 672. *Rotundo, M. (2006). [Intercorrelations between self-rated and supervisorrated dimensions of performance]. Unpublished raw data. Rotundo, M., & Spector, P. E. (2010). Counterproductive work behavior and withdrawal. In J. L. Farr & N. T. Tippins (Eds.), Handbook of employee selection (pp. 489 –511). New York, NY: Routledge. Sackett, P. R. (2002). The structure of counterproductive work behaviors: Dimensionality and relationships with facets of job performance. International Journal of Selection and Assessment, 10, 5–11. doi:10.1111/ 1468-2389.00189 Sackett, P. R., & DeVore, C. J. (2001). Counterproductive behaviors at work. In N. Anderson, D. S. Ones, H. K. Sinangil, & V. Viswesvaran

(Eds.), Handbook of industrial, work, and organizational psychology (Vol. 1, pp. 145–164). London, England: Sage. *Sady, K., Spitzmuller, C., & Witt, L. A. (2008, August). Good employee, bad business: An interactionist approach to workplace deviance. Paper presented at the annual meeting of the Academy of Management, Anaheim, CA. Salgado, J. F. (2002). The Big Five personality dimensions and counterproductive behaviors. International Journal of Selection and Assessment, 10, 117–125. doi:10.1111/1468-2389.00198 Schmidt, F. L., Viswesvaran, C., & Ones, D. S. (2000). Reliability is not validity and validity is not reliability. Personnel Psychology, 53, 901– 912. doi:10.1111/j.1744-6570.2000.tb02422.x *Skarlicki, D. P., & Folger, R. (1997). Retaliation in the workplace: The roles of distributive, procedural, and interactional justice. Journal of Applied Psychology, 82, 434 – 443. doi:10.1037/0021-9010.82.3.434 Spector, P. E. (2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9, 221–232. doi: 10.1177/1094428105284955 *Spector, P. E., Bauer, J., & Fox, S. (2010). Measurement artifacts in the assessment of counterproductive work behavior and organizational citizenship behavior: Do we know what we think we know? Journal of Applied Psychology, 95, 781–790. doi:10.1037/a0019477 Spector, P. E., Fox, S., Penney, L. M., Bruursema, K., Goh, A., & Kessler, S. (2006). The dimensionality of counterproductivity: Are all counterproductive behaviors created equal? Journal of Vocational Behavior, 68, 446 – 460. Spector, P. E., & Jex, S. M. (1998). Development of four self-report measures of job stressors and strain: Interpersonal Conflict at Work Scale, Organizational Constraints Scale, Quantitative Workload Inventory, and Physical Symptoms Inventory. Journal of Occupational Health Psychology, 3, 356 –367. doi:10.1037/1076-8998.3.4.356 *Stamper, C. L., & Masterson, S. S. (2002). Insider or outsider? How employee perceptions of insider status affect their work behavior. Journal of Organizational Behavior, 23, 875– 894. doi:10.1002/ job.175 Stewart, S. M., Bing, M. N., Davison, H. K., Woehr, D. J., & McIntyre, M. D. (2009). In the eyes of the beholder: A non-self-report measure of workplace deviance. Journal of Applied Psychology, 94, 207–215. doi: 10.1037/a0012605 *Thau, S., Crossley, C., Bennett, R. J., & Sczesny, S. (2007). The relationship between trust, attachment, and antisocial work behaviors. Human Relations, 60, 1155–1179. doi:10.1177/0018726707081658 *Tubre´, T., Arthur, W., Jr., & Bennett, W., Jr. (2006). General models of job performance: Theory and practice. In W. Bennett, C. E. Lance, & D. J. Woehr (Eds.), Performance measurement: Current perspectives and future challenges (pp. 175–203). Mahwah, NJ: Erlbaum. Van Katwyk, P. T., Fox, S., Spector, P. E., & Kelloway, E. K. (2000). Using the Job-Related Affective Well-Being Scale (JAWS) to investigate affective responses to work stressors. Journal of Occupational Health Psychology, 5, 219 –230. Vardi, Y., & Weitz, E. (2004). Misbehavior in organizations: Theory, research, and management. Mahwah, NJ: Erlbaum. Viswesvaran, C., & Ones, D. S. (2000). Measurement error in “Big Five Factors” personality assessment: Reliability generalization across studies and measures. Educational and Psychological Measurement, 60, 224–235. Viswesvaran, C., Ones, D. S., & Schmidt, F. L. (1996). Comparative analysis of the reliability of job performance ratings. Journal of Applied Psychology, 81, 557–574. Viswesvaran, C., Schmidt, F. L., & Ones, D. S. (2002). The moderating influence of job performance dimensions on convergence of supervisory and peer ratings of job performance: Unconfounding construct-level convergences and rating difficulty. Journal of Applied Psychology, 87, 345–354. doi:10.1037/0021-9010.87.2.345 Watson, D., Clark, L. A., & Tellegen, A. (1998). Development and validation of

SELF- AND OTHER-REPORTS OF CWB brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070. *White, L. A., Young, M. C., & Rumsey, M. G. (2001). Able implementation issues and related research. In J. P. Campbell & D. J. Knapp (Eds.), Exploring the limits in personnel selection and classification (pp. 525–558). Mahwah, NJ: Erlbaum.

629

Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and organizational commitment as predictors of organizational citizenship and in-role behaviors. Journal of Management, 17, 601– 617. *Williams, M. (1999). When is personality a predictor of performance? The moderating role of autonomy (Unpublished doctoral dissertation). Florida International University, Miami.

Appendix Table A1 List of Search Terms Used in the Literature Search for Primary Data Search terms Anticitizenship Bullying Counterproductive work behavior Counterproductivity Misuse of resources Misuse of time Misuse of workplace resources Noncompliant behavior Organization motivated aggression Organizational aggression Organizational deviance Organizational misbehavior Organizational retaliation behavior Personal aggression Political deviance

Poor work attendance Production deviance Property deviance Property theft Sabotage Stealing property Stealing resources Theft Undesirable work behavior Workplace aggression Workplace deviance Workplace deviant behavior Workplace gossip Workplace misconduct

(Appendices continue)

319 173 122 173

2

3a 4a 1

Colbert et al. (2003)

239

1

Colbert et al. (2004)

233

1

Chiu & Peng (2008)

121

1

187

3

Bruk-Lee & Spector (2006)

168

2

Bordia et al. (2008)

176

N

3

Sample #

Bing et al. (2007)

Reference

WE WE CWB-I CWB-I

Supervisor CWB-I

Supervisor Supervisor Supervisor Supervisor

Supervisor CWB-O

Supervisor CWB-I

— — —

—

—

—

Coworker —

—

Supervisor CWB-O Self Coworker

—

—

—

CWB-2

Supervisor CWB-I

Supervisor CWB-O

Coworker

CWB-1

Table A2 Relevant Pieces of Information for Each Primary Sample

— — —

—

—

—

(Appendices continue)

— — —

—

—

—

— — 48.0

—

—

70.0

— — —

—

—

—

2 —

76.7

—

—

—

# of anonymity safeguards

.30 ⫺.11 — —

58.0

48.2

—

% female

—

—

—

—

d

—

—

—

—

—

r

SelfreportOtherreport

Gender Age Tenure Negative emotion Coworker conflictc Supervisor conflictc Coworker conflictb Supervisor conflictb Gender Age Gender Age Emotional Stability Conscientiousness Emotional Stability Conscientiousness Agreeableness Agreeableness Gender Age Emotional Stability Extraversion Openness Conscientiousness

OCB-Ib OCB-Ob Gender Age Tenure Gender Age Tenure Gender Age Tenure

CWB correlate

⫺.18 ⫺.18 ⫺.03 .25 .39 .14 .50 .36 ⫺.09 ⫺.07 ⫺.01 ⫺.11 ⫺.02 ⫺.23 .01 ⫺.21 ⫺.50 ⫺.55 ⫺.31 ⫺.08 ⫺.27 ⫺.20 ⫺.14 ⫺.48

⫺.29 ⫺.56 .06 ⫺.02 ⫺.02 ⫺.06 ⫺.10 ⫺.04 ⫺.03 .00 .00

Other-report-CWBcorrelate r

630 BERRY, CARPENTER, AND BARRATT

1

3

1

Ferris (2008)

Ferris et al. (2008)

Fox et al. (2007)

122

3

132

164

164

54 321 191

229

2

1 1 1

N

Sample #

de Jonge & Peeters (2009) Fallon et al. (2000) Ferris (2007)

Reference

Table A2 (continued)

—

Peer

Self CWB-O

.13

.30 .47

—

—

.49 —

.40 .55 .58

.46 — .42 .45

—

—

r

.24

—

—

—

.28 —

—

—

d

—

60.0

60.0

— —

68.0

42.0

% female

(Appendices continue)

Coworker Coworker CWB-I Coworker CWB-O —

—

Peer CWB-O

Self Self CWB-I

—

Coworker Coworker CWB-I Coworker CWB-O

Coworker — Coworker Coworker CWB-I Coworker CWB-O

—

—

CWB-2

Self CWB-O Peer CWB-I

Self CWB-O Self Self CWB-I

Self Supervisor CWB-O Self Self CWB-I

Supervisor CWB-I

Supervisor CWB-I

CWB-1

SelfreportOtherreport

—

—

—

2 —

—

—

# of anonymity safeguards

Gender Age OCB-I OCB-O Gender Age OCB-I OCB-O OCB coworker OCB-I coworker OCB-O coworker

Conscientiousness

Gender Age Emotional Stability Extraversion Openness Agreeableness Conscientiousness Gender Age Emotional Stability Extraversion Openness Conscientiousness

CWB correlate

⫺.14 ⫺.20 ⫺.17 ⫺.45 ⫺.07 .00 ⫺.23 ⫺.59 ⫺.46 ⫺.22 ⫺.57

.09

⫺.05 ⫺.09 ⫺.08 .08 ⫺.08 ⫺.37 .01 ⫺.34 .16 .04 .18 ⫺.01 .06

Other-report-CWBcorrelate r

SELF- AND OTHER-REPORTS OF CWB

631

1

1 1

1

Greenbaum et al. (2009) Hung et al. (2009)

Jones (2004)

Sample #

Goh (2007)

Reference

Table A2 (continued)

173

122 184

146

N

Self CWB-O

Self CWB-O Supervisor Self Coworker CWB-I

Supervisor Self Self CWB-I

Self Coworker

Coworker CWB-O

Coworker CWB-I

CWB-1

.17

.19 — .16 .23

— .24 .27

.42 —

—

—

r

— .45

— .00

.17 —

—

—

d

— —

— —

—

—

—

% female

(Appendices continue)

Coworker Coworker CWB-I Coworker CWB-O —

— Supervisor Supervisor CWB-I Supervisor CWB-O —

Coworker —

—

CWB-2

SelfreportOtherreport

— 1

— 1

2 —

—

—

# of anonymity safeguards

Negative affectivity

OCB

Informational justice Interpersonal justice Interpersonal conflictc Interpersonal conflictb Interpersonal constraintsc Interpersonal constraintsb Job contexts constraintsc Job contexts constraintsb Organizational constraintsc Organizational constraintsb Negative affectivity

Organizational constraintsc Interpersonal conflictc Procedural justicec Distributive justicec Organizational constraintsb Interpersonal conflictb Organizational constraintsc Interpersonal conflictc Procedural justicec Distributive justicec Organizational constraintsb Interpersonal conflictb

CWB correlate

.04

⫺.47

⫺.17 ⫺.20 .37 .50 .27 .39 .25 .38 .28 .41 .24

.12 .25 ⫺.29 ⫺.06 .36 .61 .13 .20 ⫺.35 ⫺.11 .40 .52

Other-report-CWBcorrelate r

632 BERRY, CARPENTER, AND BARRATT

1

1a 1b

1

1

1 1

Judge et al. (2006b)

Kessler et al. (2009)

Kickul et al. (2001)

Kidwell & Bennett (2001)

Kuenzi & Carson (2009) Lee (2000) 311 146

556

165

96 85

64

131

158

2

2

N

Sample #

Judge et al. (2006a)

Reference

Table A2 (continued)

Supervisor CWB-I Self Coworker

Self Supervisor

Self Supervisor Supervisor Supervisor

Supervisor

Self Supervisor

Coworker

Coworker

CWB-1

— Coworker —

Supervisor —

Supervisor — — —

—

Supervisor —

—

CWB-2

— .32 —

.25 — — —

—

.94 —

—

—

d

— —

— — 71.8 49.7

—

— —

82.0

—

% female

(Appendices continue)

— .33 —

.15 —

.72 — — —

—

.20 —

—

—

r

SelfreportOtherreport

— 1

0

—

—

—

1

—

—

# of anonymity safeguards

OCB-Ia OCB (composite)c OCB-Oc OCB-Ic

OCB-I (supervisor)

Gender Tenure Procedural justice Interactional justice

Neuroticism Extraversion Openness Agreeableness Conscientiousness OCB-self Interpersonal justice Job satisfaction Job satisfaction Gender

Procedural justice–Pay Procedural justice–Direct Procedural justice–Scheduling Distributive justice–Pay Distributive justice–Scheduling OCBb OCB-Ob Interpersonal justice Informational justice Direct interactional justice OCBb Gender Age

CWB correlate

⫺.33 ⫺.07 ⫺.01 ⫺.11

⫺.09

⫺.15 ⫺.26 ⫺.49 ⫺.57

⫺.40 ⫺.34 ⫺.18 ⫺.13 ⫺.11 ⫺.40 ⫺.02 ⫺.05 .14 .12 ⫺.10 ⫺.05 ⫺.07 ⫺.14 ⫺.24 ⫺.14 ⫺.26 ⫺.24

⫺.02 .08

⫺.12 ⫺.09 ⫺.23

Other-report-CWBcorrelate r

SELF- AND OTHER-REPORTS OF CWB

633

Sample #

1

1

1

1

1

1

1

Reference

Lee & Allen (2002)

Mount et al. (2006)

O’Brien (2009)

O’Brien & Allen (2007)

O’Brien & Loch (2005)

Penney & Spector (2005)

Puffer (1987)

Table A2 (continued)

141

155

176 176 175 172 111

207

176

141

155

N

Supervisor

Self Self CWB-I Self CWB-O Coworker

Self Coworker

Self Self CWB-I Self CWB-O Supervisor

Coworker

Self Self CWB-I Self CWB-O Supervisor

Coworker

CWB-1

—

—

—

—

—

.24 ⫺.05 .30 — .22 —

—

—

—

—

% female

—

.41 —

—

.02 —

—

d

.25 —

.22 —

.36 .48 .19 —

—

.38 .48 .21 —

—

r

(Appendices continue)

Coworker Coworker CWB-I Coworker CWB-O —

Coworker —

Supervisor Supervisor CWB-I Supervisor CWB-O —

—

Supervisor Supervisor CWB-I Supervisor CWB-O —

—

CWB-2

SelfreportOtherreport

—

2

—

1

—

—

# of anonymity safeguards Age Tenure Negative affect Procedural justice Emotional Stability Extraversion Openness Agreeableness Conscientiousness Job satisfaction Extraversion Conscientiousness Emotional Stability Agreeableness OCB-Ia OCB-Oa OCB-Ic OCB-Oc Org. justice Job satisfaction Conscientiousness OCBc Job satisfaction Distributive justice Procedural justice OCBb Tenure Negative affect Conflictc Conflictb Constraints Job satisfaction OCBa Job satisfaction

CWB correlate

.00 .10 .14 ⫺.05 ⫺.22 ⫺.11 ⫺.22 ⫺.14 ⫺.23 ⫺.19 ⫺.04 ⫺.18 ⫺.02 ⫺.09 ⫺.48 ⫺.54 ⫺.16 ⫺.14 ⫺.15 ⫺.10 ⫺.16 ⫺.12 ⫺.12 .13 ⫺.08 ⫺.48 ⫺.04 .19 .22 .50 .07 ⫺.25 ⫺.74 ⫺.16

Other-report-CWBcorrelate r

634 BERRY, CARPENTER, AND BARRATT

1

1 1

1

1 2 1 2

1

2 3

1 1

Rotundo (2006) Sady et al. (2008)

Skarlicki & Folger (1997)

Spector et al. (2010)

Stamper & Masterson (2002)

Thau et al. (2007)

Tubre´ et al. (2006) White et al. (2001)

Sample #

Richards & Schat (2011)

Reference

Table A2 (continued)

287 590

87 106

257

119 140 119 140

167

30 171

153 153

N

Self Supervisor

Supervisor Supervisor

Supervisor

Self Self Supervisor Supervisor

Coworker

Self Self CWB-I Self CWB-O Self Coworker

CWB-1

—

Supervisor —

–

—

Supervisor Supervisor — —

—

.14 —

— —

—

.48 .42 — —

—

.72 .70 .67 .17 —

r

.47 —

— —

—

.48 .61 — —

—

—

.19

d

— —

53.0 —

75.0

— —

—

— —

% female

(Appendices continue)

Supervisor Supervisor CWB-I Supervisor CWB-O Supervisor —

CWB-2

SelfreportOtherreport

1 —

— —

—

2 2

—

— —

# of anonymity safeguards

Emotional Stability Agreeableness Conscientiousness

Emotional Stability Satisfaction with work Satisfaction with coworkers Distributive justice Procedural justice Interactional justice OCBc OCB-I compositec OCB-O compositec Interpersonal conflict at work OCBc OCB-I compositec OCB-O compositec Interpersonal conflict at work Tenure Gender Age OCB-Ia Gender Age Tenure Age Tenure

CWB correlate

.03 .07 .20

⫺.11 ⫺.19 ⫺.17 ⫺.44 ⫺.53 ⫺.54 ⫺.33 ⫺.32 ⫺.26 .39 ⫺.20 ⫺.57 ⫺.02 .37 ⫺.16 .21 ⫺.06 ⫺.45 ⫺.04 .00 .08 ⫺.03 .07

Other-report-CWBcorrelate r

SELF- AND OTHER-REPORTS OF CWB

635

1

M. Williams (1999)

94

N Supervisor

CWB-1 —

CWB-2 —

r —

d

SelfreportOtherreport

—

% female —

# of anonymity safeguards

OCBa Conscientiousness Neuroticism Extraversion Openness Agreeableness Job satisfaction

CWB correlate

⫺.35 ⫺.06 ⫺.12 .21 ⫺.05 ⫺.04 ⫺.05

Other-report-CWBcorrelate r

Received June 9, 2010 Revision received November 11, 2011 Accepted November 28, 2011 䡲

Note. CWB-1 ⫽ first CWB variable (e.g., self-ratings of CWB); CWB-2 ⫽ second CWB variable (e.g., supervisor ratings of CWB); Self-report-Other-report r ⫽ correlation between self- and other-report CWB; Self-report-Other-report d ⫽ standardized mean difference between self- and other-ratings of CWB (positive ds mean self-report was higher); % female ⫽ percentage of the sample that was female; # of anonymity safeguards ⫽ the number of assurances of anonymity provided to participants (0, 1, or 2); CWB correlate ⫽ variable correlated with other-report CWB; Other-report-CWB-correlate r ⫽ correlation between other-report CWB and the CWB correlate; Self ⫽ self-report; Supervisor ⫽ supervisor report; Coworker ⫽ coworker report; OCB-I and OCB-O ⫽ Interpersonal- and organizational target organizational citizenship behaviors, respectively; CWB-I and CWB-O ⫽ interpersonal- and organizational target counterproductive work behaviors, respectively; WE ⫽ withholding effort. Dashes respresent instances in which that piece of information was not available. a Supervisor-ratings. b Coworker ratings. c Self; all other nonbehavioral correlates were self-reported.

Sample #

Reference

Table A2 (continued)

636 BERRY, CARPENTER, AND BARRATT