j.2044-8325.2011.02032.x

Upload: jonathan-james-mercado

Post on 14-Oct-2015

9 views

Category:

Documents


0 download

DESCRIPTION

Inudstrial Psychology Journal

TRANSCRIPT

  • 5/24/2018 j.2044-8325.2011.02032.x

    1/22

    199

    Journal of Occupational and Organizational Psychology (2012), 85, 199220C2011 The British Psychological Society

    The

    British

    Psychological

    Society

    www.wileyonlinelibrary.com

    The deviant citizen: Measuring potential positiverelations between counterproductive workbehaviour and organizational citizenshipbehaviour

    Suzy Fox1, Paul E. Spector2, Angeline Goh3, Kari Bruursema4and Stacey R. Kessler5

    1Institute of Human Resources and Employment Relations, Loyola UniversityChicago, Illinois, USA

    2Department of Psychology, University of South Florida, Florida, USA3Permanente Medical Group, California, USA4Department of Psychology, Montclair State University, New Jersey, USA5Department of Management and Information Sciences, Montclair State University,New Jersey, USA

    Studies have shown a strong negative correlation between counterproductive work

    behaviour (CWB) and organizational citizenship behaviour (OCB), and opposite cor-

    relations with hypothesized antecedents. Such observed correlations may have been

    erroneously caused by three measurement artefacts: items measuring absence of

    CWBs, rather than behaviours that exceed requirements or expectations in OCB

    scales; supervisory halo; and agreement rather than frequency response format. A new

    OCB scale, the OCB-checklist (OCB-C) was used that did not have these artefacts.

    Contrary to prior expectations from the literature, positive relations were found

    between CWB and OCB, and stressors and OCB. Theoretical explanations for positive

    CWB/OCB relations (demand-elicited OCB, social loafing, work process problems,rater perceptions and attributions, and aggravated job stress processes) are discussed.

    Counterproductive work behaviour (CWB; behaviour that harms) and organizational

    citizenship behaviour (OCB; behaviour that helps) are forms of extra-task behaviourthat are often considered opposites (e.g., Sackett, 2002). Indeed studies have shownan inverse and often strong relationship between the two (e.g., Dineen, Lewicki, &Tomlinson, 2006; Lee & Allen, 2002; Rotundo & Sackett, 2002), and they often correlate

    in opposite directions with potential antecedents, such as justice (Cohen-Charash &Spector, 2001; Colquitt, Conlon, Wesson, Porter, & Ng, 2001), job satisfaction (Dalal,

    Correspondence should be addressed to Suzy Fox, Institute of Human Resources and Employment Relations, Loyola UniversityChicago, 820 N. Michigan Avenue, Chicago, IL 60611 (e-mail:[email protected]).

    DOI:10.1111/j.2044-8325.2011.02032.x

  • 5/24/2018 j.2044-8325.2011.02032.x

    2/22

    200 Suzy Fox et al.

    2005), and interpersonal conflict (Fox, Spector, & Miles, 2001; Zellars, Tepper, & Duffy,2002). This suggests that factors leading to high levels of one lead to low levels of theother (Spector & Fox, 2002). A few studies, however (Kelloway, Loughlin, Barling, &Nault, 2002; Miles, Borman, Spector, & Fox, 2002), failed to find these expected results

    when they selectively retained from existing scales only those items they judged to

    be non-overlapping between CWB and OCB, but it was Dalals (2005) meta-analysisthat posed distinct methodological challenges by identifying methodological artefactsthat might have distorted relationships between CWB and OCB. Recent work has

    investigated Dalals methodological artefacts, and raised substantive issues that challengethe generally assumed negative association between CWB and OCB (Spector, Bauer, &Fox, 2010; Spector & Fox, 2010a, 2010b).

    The first purpose of this paper is to introduce the OCB-checklist (OCB-C), a new

    measure that addresses methodological artefacts that may have impacted and distortedour conclusions about CWBOCB relations. Second, two studies are presented todemonstrate that, when we used a measure that avoided those methodological artefacts,

    we found that CWB and OCB can be positively related, and they both can relate similarlyto antecedent variables. We offer suggestions (to be pursued in future research) ofsubstantive conditions, under which acts of OCB can actually lead to CWB and acts ofCWB can lead to OCB. Dalal (2005) and Spectoret al. (2010) have already shown thatcontrolling measurement artefacts in OCB results in quite different correlations with

    CWB. What kind of picture might emerge in the relationship of an artefact-free OCBscale with other potential antecedents of this form of behaviour?

    We propose that in some cases, parallel and opposite processes will still occur, aspredicted by integrative models of voluntary or extra-task work behaviour (e.g., Spector

    & Fox, 2002). Expectations drawn from this framework as well as the literature (e.g.,

    Dineen et al., 2006; Lee & Allen, 2002) are that CWB and OCB would be negativelycorrelated, and that antecedents such as job stressors and job-related emotions wouldbe related to CWB and OCB in opposite directions. However, in other cases, situational

    events and evaluative/affective processes might elicit the direct interplay of CWB andOCB by the same actor, in response to the same antecedents. The deconstructionof this process (i.e., under what circumstances will the one framework and not theother predominate?) is beyond the scope of the current study. Our objective here is to

    demonstrate that using methodologically appropriate measures, the negative CWB/OCBrelations will not occur, and, in fact, certain situational variables will co-occur with bothCWB and OCB.

    Defining the domain of CWB and OCB

    CWB has been studied by a number of researchers from a variety of theoretical perspec-

    tives including aggression (Douglas & Martinko, 2001; Fox & Spector, 1999; Neuman &Baron, 1998; OLeary-Kelly, Griffin, & Glew, 1996), deviance (Hollinger, 1986; Robinson& Bennett, 1995), retaliation (Skarlicki & Folger, 1997), and revenge (Bies, Tripp, &Kramer, 1997).What these approaches have in common is the perspective that thesebehaviours are volitional (as opposed to accidental or mandated) and harm organizations

    and/or organization stakeholders, such as clients, co-workers, customers, and supervisors(Spector & Fox, 2005). Researchers commonly categorize CWB according to the targetof the behaviour, that is, acts directed against the organization itself versus other persons

    in the organization (Bennett & Robinson, 2000; Foxet al., 2001).

  • 5/24/2018 j.2044-8325.2011.02032.x

    3/22

    The deviant citizen 201

    In contrast, efforts to define OCB have seen a long period of controversy, withlittle consensus over the parameters and even conceptualization and evaluation of theconstruct. Although there have been disagreements over definitional issues such as

    whether or not OCB is extra role and whether or not it is rewarded (Organ, 1997), to

    the best of our knowledge no published sources define OCB as theabsenceof CWB. We

    conceptualize OCB as helpful behaviours that support the social fabric of the organizationthat are outside of the core job tasks (Organ, 1997). Many efforts have been made toconsider relationships between OCB and CWB. By clearly distinguishing these extra-

    task behaviours as acts that harm (CWB), and acts that help (OCB) organizations andtheir members, the stage was set early on (Spector & Fox, 2002) for the expectationthat these would be parallel or opposite forms of behaviour, and that CWB and OCB

    would relate oppositely with antecedents and consequences. Indeed, some research,

    using common measures of CWB (e.g., Bennett & Robinson, 2000; Fox et al., 2001)and OCB scales (e.g., Podsakoff, MacKenzie, Moorman, & Fetter, 1990; Smith, Organ, &Near, 1983; Williams & Anderson, 1991) demonstrated such inverse relationships (e.g.,

    Rotundo & Sackett, 2002). Yet, a closer look at the measures used in these studies hasprompted a re-evaluation of the apparent opposite nature of CWB and OCB.

    OCB measurement issues to be addressed

    Dalal (2005) conducted a meta-analysis of studies that assessed both forms of behaviour,and concluded that whereas CWB and OCB are strongly and negatively related acrossmost studies, the relationship between CWB and OCB is substantially affected bymethodological artefacts. Dalal discussed three methodological artefacts that are likely to

    distort assessment of relations between CWB and OCB: item overlap between measures

    of both CWB and OCB, whether an agreement (does the person do this behaviour?) orfrequency (how often does the person do this behaviour?) response is used, and whetherself-reports or supervisor reports are utilized. The magnitude of correlation was only

    strong when there were overlapping items between scales (which Dalal calls antitheticalitems, i.e., the same or similar item, reverse worded or reverse coded, appears in boththe CWB and OCB scales), the scales asked for agreement rather than frequency, orsupervisor reports were compared. He suggested that the observed relationships may

    have been inflated due to measurement problems and halo in supervisor ratings. In afollow-up study, Spectoret al.(2010) were able to isolate the three individual artefactsthat were confounded in the studies Dalal (2005) had in his meta-analysis. Their results

    were consistent with Dalal in showing that CWB and OCB had either no relationship,or a slight positive relationship when various combinations of the three artefacts wereremoved.

    In Study 1, we will describe the development of the organizational citizenshipchecklist (OCB-C), which was designed to avoid these measurement artefacts. Of the

    three, overlapping or antithetical item content had the biggest effect on CWBOCBcorrelations. Table 1 illustrates several examples of items from three prevalently usedOCB scales (Podsakoffet al., 1990; Smith, Organ, & Near, 1983; Williams & Anderson,1991). Antithetical items reflect similar behaviours that appear in both the OCB andCWB scales, often reverse worded and reverse coded. For example, all three OCB scales

    include Takes undeserved breaks or Does not take extra breaks; compare this to theCWB checklist item (CWB-C, Spectoret al., 2006) Taken a longer break than you wereallowed to take, or the Bennett and Robinson (2000) deviance scale item Taken an

    additional or longer break than is acceptable at your workplace. The Podsakoffet al.

  • 5/24/2018 j.2044-8325.2011.02032.x

    4/22

    202 Suzy Fox et al.

    Table 1. Antithetical OCB/CWB items

    Scale OCB item CWB item from CWB-checklist

    SON Takes undeserved breaks (R) Taken a longer break than you were allowed to take

    WA Takes undeserved work breaks (R) Left work earlier than you were allowed to

    SON Coasts towards the end of the day (R) Tried to look busy while doing nothingSON and

    WA

    Great deal of time spent with

    personal phone conversations (R)

    Daydreamed rather than did your work

    PMMF Obeys company rules and regulations

    even when no one is watching

    Purposely failed to follow instructions

    PMMF Consumes a lot of time complaining

    about trivial matters (R)

    Complained about insignificant things at work

    WA Complains about insignificant things

    at work (R)

    WA Conserves and protects

    organizational property

    Purposely wasted your employers materials/

    suppliesPurposely damaged a piece of equipment or

    property

    PMMF Does not abuse the rights of others Ignored someone at work

    Refused to help someone at work

    Purposely interfered with someone at work doing

    his/her job

    Insulted someone about their job performance

    Made fun of someones personal life

    Started an argument with someone at work

    Verbally abused someone at work

    Made an obscene gesture (the finger) to someoneat work

    Threatened someone at work with violence

    Threatened someone at work, but not physically

    Hit or pushed someone at work

    Insulted or made fun of someone at work

    Note. PMMF, Podsakoff, MacKenzie, Moorman, and Fetter (1990); SON, Smith, Organ, and Near (1983);

    WA, Williams and Anderson (1991).

    (1990) item, Does not abuse the rights of others is similar to several CWB-C itemscovering starting rumours, being nasty, insulting, threatening, and making fun of co-workers and clients. In some cases, CWB and OCB items represent opposite ends ofthe same construct. For example, Williams and Anderson (1991) include Conservesand protects organizational property, while the CWB-C includes Purposely wasted

    your employers materials/supplies and Purposely damaged a piece of equipment or

    property. Likewise, the Skarlicki and Folger (1997) retaliation scale contains the itemOn purpose damaged equipment or work process.

    Dalal (2005) focused on the strength of negative relationships between CWB andOCB; however, a few studies have even found positive relations or relations in the same

    direction with variables assumed to relate oppositely with CWB and OCB (see Marcus,Schuler, Quell, & Humpfner, 2002). Mileset al.(2002) omitted overlapping items fromtheir OCB scale based on the Coleman and Borman (2000) Citizenship Performance Scale,and found not only non-significant CWB/OCB relations, but also significant relations

  • 5/24/2018 j.2044-8325.2011.02032.x

    5/22

    The deviant citizen 203

    in the same direction of CWB and OCB with organizational constraints, interpersonalconflict, and workload. Venkataramani and Dalal (2007) also found positive correlationsbetween interpersonal helping and harming in a non-work setting. These studies suggestthat at least some of the findings in the literature concerning CWB and OCB were affected

    by methodological issues and not due to CWB and OCB being exclusively opposite forms

    of behaviour.

    Substantive challenges to assumptions that CWB and OCB are opposite forms

    of behaviour

    Some researchers (e.g., Bies & Tripp, 2005; Duffy, Ganster, & Pagon, 2002; Fox, 2005,2008; Galparin & Burke, 2006; Mitchell & Bennett, 2005; Skarlicki & Folger, 1997;Spector & Fox, 2010a, 2010b; Warren, 2003) have developed substantive and theoretical

    challenges to the assumption that employees who are more likely to engage in OCBs(citizens), are less likely to perform CWBs (deviants), and vice versa. We are notmaintaining that mechanisms do not exist that might underlie both CWB and OCB

    in opposite ways. However, sometimes the situation seems to strongly call for OCB,which Spector and Fox (2010a; 2010b) called demand-elicited OCB and Vigoda-Gadot(2006) called compulsory citizenship behaviour a different dynamic than voluntarybeneficence.

    Substantive drivers of positive CWB/OCB relations may include co-worker lack of

    performance and social loafing, particularly when the co-workers poor performanceis attributed to internal and controllable factors, such as poor job attitudes (e.g.,commitment), or lack of effort (e.g., social loafing). In that case, the employee maystill feel compelled to contribute OCB to complete the co-workers tasks, but may also

    respond with stress perceptions, negative emotions and attitudes, and ultimately CWB,

    often directed towards the very co-worker the employee is helping. Interpersonal conflictmay be both an antecedent to, and a consequence of, perceived imbalances in levels ofcontributions to work tasks and goals, and may result in efforts to somehow right the

    imbalance.We suggest that conditions at work often create necessities for engaging in extra-

    task behaviour in order to remain productive on the job. Organizational constraintsare conditions at work that interfere with performing job tasks (Peters & OConnor,

    1980). Constraints can lead to an employee going beyond assigned tasks, involvingsuch activities as informally organizing co-workers to support efforts to get tasks done,finding ways to learn new skills not required, or taking on tasks not part of the job,

    which are clearly forms of OCB. However, research has consistently found organizationalconstraints to be associated with anger and frustration that often lead to CWB (Fox et al.,2001). Indeed organizational constraints have been shown to relate positively to bothCWB and OCB (Mileset al., 2002).

    In general, to the extent that engaging in additional work precipitates negative

    outcomes such as overload, injustice perceptions, interpersonal conflict, and escalatingcitizenship requirements (Bolino, Turnley, & Niehoff, 2004), OCB may actually increase

    job stress. Given the robust linkage in the research literature between job stress and CWBas a form of behavioural strain (e.g., Hershcovis et al., 2007), this may be a key pathlinking some forms of OCB with CWB. These examples underscore the point that there

    are common antecedents that can lead to both CWB and OCB, often through a chain ofevents from one behaviour to the other. In the current study, we consider organizationalconstraints and interpersonal conflict as potential situational variables that may give rise

    to an employee engaging in high levels of both CWB and OCB.

  • 5/24/2018 j.2044-8325.2011.02032.x

    6/22

    204 Suzy Fox et al.

    The current studies

    The above discussion of substantive and attributional factors contributing to positiveOCB/CWB relations in some situations leads us to the expectation that, by eliminatingthe methodological issues reported by Dalal, the negative OCB/CWB relations, commonly

    found with prior (i.e., overlapping) measures, will disappear. While a previous paper

    (Spectoret al., 2010) presents an experimental study that manipulated combinations ofall three of Dalals (2005) measurement concerns, the current paper presents two fieldsurvey tests of OCB/CWB relations, using a measure of OCB specifically developed to

    eliminate antithetical item overlap and agreement responses.

    STUDY 1

    The purpose of Study 1 was to challenge the negative relations (and test the distinctive-ness) of CWB and OCB, using the OCB-C designed to remove potential methodological

    confounds described by Dalal (2005). The initial roadblock to conducting this researchwas locating a scale to assess OCB that did not contain the problems of item overlap andagree format.

    First, all measures of OCB we could find contained at least some items that reflectedthe absence of harmful behaviours rather than just helpful behaviours (Table 1). Second,

    we wanted to be sure that the items were all clearly behavioural in nature and that itwould make sense to ask subjects to make frequency judgements. The B in OCB refersto behaviour, and thus scales to assess it should ask about behaviour itself and not otherthings. Some of the items of the existing scales involve beliefs or attributions about

    the persons behaviour and/or motives that seem more reflective of personality than

    behaviour itself. For example, the item from the Podsakoffet al. (1990) scale Alwaysfocuses on whats wrong, rather than the positive side seems like emotional stability,and the item Believes in giving an honest days work for an honest days pay seems

    like conscientiousness. The accuracy and relevance of attributions of intention havebeen challenged in the extra-role literature (Bolinoet al., 2004; Van Dyne, Cummings, &McLean Parks, 1995), particularly when the respondent is rating the behaviour of others(e.g., supervisor, subordinate, or peer). Therefore, a more clear-cut behavioural checklist

    asking about what the person has actually done would be preferable to items asking oneorganization member to make judgements about attributions, beliefs, or personality ofanother.

    Therefore, we developed the OCB-C in a format similar to that of the CWB-C, basedupon actual behaviours or incidents submitted by subject matter experts. Specifically,38 employed graduate students and alumni of Masters of Science in Human Resources,Industrial/Organizational Psychology, and MBA programmes were instructed to generateas many examples as they could remember of themselves or people with whom they

    worked engaging in OCBs in their workplace.A total of 214 critical incidents were generated. After redundancies were eliminated,

    like items were consolidated, and unusable suggestions were eliminated (helped save acoworker from eternal damnation), we were left with a 36-item checklist. Interestingly,

    while we had initially been concerned about how we would develop a process to

    recognize and eliminate items that overlapped with CWB items, not a single instanceof an antithetical CWB incident was submitted. In a preliminary study, we furtherreduced the 36 items by eliminating low base-rate behaviours (high frequency of neverresponses), resulting in a 20-item OCB-C. Items include Offered suggestions to improve

  • 5/24/2018 j.2044-8325.2011.02032.x

    7/22

    The deviant citizen 205

    Table 2. OCB-checklist with Study 1 means (n = 169)

    Mean score

    (range: 15)

    1. Picked up meal for others at work. 2.32

    2. Took time to advise, coach, or mentor a co-worker. 2.893. Helped co-worker learn new skills or shared job knowledge. 3.30

    4. Helped new employees get oriented to the job. 2.84

    5. Lent a compassionate ear when someone had a work problem. 3.45

    6. L ent a compassionate ear when someone had a personal problem. 3.26

    7. Changed vacation schedule, workdays, or shifts to accommodate co-workers needs. 1.95

    8. Offered suggestions to improve how work is done. 2.79

    9. Offered suggestions for improving the work environment. 2.71

    10. Finished something for co-worker who had to leave early. 2.46

    11. Helped a less capable co-worker lift a heavy box or other object. 2.33

    12. Helped a co-worker who had too much to do. 2.61

    13. Volunteered for extra work assignments. 2.4314. Took phone messages for absent or busy co-worker. 2.99

    15. Said good things about your employer in front of others. 2.77

    16. Gave up meal and other breaks to complete work. 2.96

    17. Volunteered to help a co-worker deal with a difficult customer, vendor, or

    co-worker.

    2.18

    18. Went out of the way to give co-worker encouragement or express appreciation. 2.78

    19. Decorated, straightened up, or otherwise beautified common work space. 2.89

    20. Defended a co-worker who was being put-down or spoken ill of by other

    co-workers or supervisor.

    2.23

    how work is done, Volunteered for extra work assignments, and Lent a compassionate

    ear when someone had a work problem. Corresponding with the CWB-C format, eachitem on the OCB-C asks the employee to indicate how often the target person (selfor other) has done each of the behaviours on the present job. The five responsechoices range from 1 = neverto 5 = every day. Table 2 presents the resulting 20-item

    OCB-C.In summary, the new OCB-C has the advantages that it is behavioural (i.e., does not

    ask respondents to make personality-based judgements for employee behaviours), asks

    for ratings of behaviour frequency rather than agreement, focuses on behaviours thatemployees themselves view as above and beyond role requirements, and does not reflectthe mere absence of CWBs.

    In addition to the CWB and OCB-Cs, we choose six variables organizationalconstraints, interpersonal conflict, job satisfaction, negative emotions, positive emotions,and justice that are commonly included in models of CWB/OCB and have been studied

    as antecedents of both CWB (e.g., Cohen-Charash & Spector, 2001; Fox, Spector, Goh, &Bruursema, 2007) and OCB (e.g., Cropanzano, Bowen, & Gilliland, 2007; Organ, 1997;Organ & Ryan, 1995).

    We surveyed both employees and their co-workers. Whereas the Spectoret al. (2010)

    study included employee and supervisor ratings, it may be argued that many discretionarywork behaviours that are hidden from the supervisor are observed by co-workers.Likewise, co-workers may be more aware of workplace constraints and conflicts thantheir supervisors. Thus, we chose to collect data on constraints, interpersonal conflict,

  • 5/24/2018 j.2044-8325.2011.02032.x

    8/22

    206 Suzy Fox et al.

    CWB, and OCB from both the focal employees and their co-workers. Affect, satisfaction,and justice perceptions were rated only by the focal employees.

    We based our first set of hypotheses on the wide array of substantive explanations,presented above, of circumstances in which the same individual might engage in both

    CWB and OCB in response to workplace demands and conditions. These include super-

    visory demands for additional performance, co-worker lack of performance, resourceconstraints, and job stress and attributional sense making in response to demand-elicitedOCB. These expectations are further supported by the Miles et al. (2002) findings of

    positive relationships of these stressors with both CWB and OCB.

    Hypothesis 1: CWB regardless of rating source will be positively related to OCB.

    Hypothesis 2: Employee-rated organizational constraints, conflict, and negative emo-tions will be positively related to employee-rated and co-worker rated

    CWB and OCB.

    Hypothesis 3: Co-worker rated organizational constraints and conflict will be positively

    related to employee-rated and co-worker rated CWB and OCB.

    Hypothesis 4 is consistent with traditional findings of attitudinal and affective relations

    with CWB and OCB. Prior research and theory present no substantial contradictions tothese commonly reported relations.

    Hypothesis 4: Employee-rated job satisfaction, positive emotions, distributive, andprocedural justice will be negatively related to employee-rated or co-

    worker rated CWB and positively related to OCB.

    Method

    Participants and procedure

    The research team distributed surveys in five organizations in Chicago and Tampa. Inaddition, students enrolled in a Master of Science in Human Resources programme

    in Chicago, all full-time employees and mostly managers, distributed the surveys toemployees in their respective organizations. In each case, two survey booklets weregiven to each employee, a self-report (Employee) form and a separate (Coworker)

    form, with instructions to give the co-worker form to a peer familiar with the employeeswork situation and behaviour. In all, 169 employee surveys were returned, for which 136matching surveys from co-workers were also returned. Of the employees who responded,24.4% were men and 75.6% were women. Of the co-workers who responded, 21.6% weremen and 78.4% were women. However, only 56.2% of the employeeco-worker dyads

    were of the same gender. Of the employees, 26.4% were managers and 73.6% werenon-managerial employees; 91.3% classified themselves as white-collar workers.

    Measures

    Organizational constraints were assessed with the Organizational Constraints Scale (OCS;

    Spector & Jex, 1998), an 11-item scale based on constraint areas identified by Petersand OConnor (1980). It measures the frequency with which employees encounterbarriers to job performance, such as rules and procedures, availability of resources, co-

    workers, interruptions, and inadequate training. Interpersonal conflict was measured

  • 5/24/2018 j.2044-8325.2011.02032.x

    9/22

    The deviant citizen 207

    with the four-item Interpersonal Conflict at Work Scale (ICAWS; Spector & Jex, 1998)that measures the frequency with which the employee experienced arguments, yelling,and rudeness in interactions with people at work. Response choices for both scalesrange from 1 = neverto 5 = every day.

    Distributive and procedural justice were assessed with scales reported in Moorman

    (1991). The six-item Distributive Justice Index, originally developed by Price and Mueller(1986), measures the employees perception of being fairly rewarded considering workinputs. There were five response choices ranging from 1 = very unfairly to 5 = veryfairly. The 12-item procedural justice scale includes both interactional and proceduraljustice items (Moorman, Blakely, & Niehoff, 1998). Response choices ranged from 1 =

    strongly disagreeto 7 = strongly agree.Job satisfaction was measured with a three-item scale derived from the Michigan

    Organizational Assessment Scale (Cammann, Fishman, Jenkins, & Klesh, 1979), withresponse choices ranging from 1 = disagree very muchto 6 = agree very much.

    Negative and positive emotions were measured with the Job-Related Affective Well-

    Being Scale (JAWS), developed by Van Katwyk, Fox, Spector, and Kelloway (2000). The20-item JAWS measures a wide range of emotions experienced in response to the job.Each item on the JAWS asks employees to indicate how often any part of the present

    job has made them feel a particular emotion (e.g., anxious, enthusiastic, or furious). Thefive response choices range from 1 = almost never to 5 = extremely often or always.

    A positive emotions score was obtained by summing the scores on the 10 positiveaffect items, with high scores representing high levels of positive emotion on the job.

    A negative emotions score was obtained by summing scores on the 10 negative affectitems, with high scores representing high levels of negative emotion on the job.

    CWB was assessed with the 45-item CWB-C (Spector et al., 2006). Each item asks

    the employee to indicate how often he or she has done each of the behaviours on theirpresent job. The five response choices range from 1 = neverto 5 = every day.

    OCB was measured with the OCB-C, as described above.

    Results and Discussion

    Table 2 presents the mean ratings on the 20 items of the OCB-C. Table 3 contains means,standard deviations, correlations, and internal consistencies (coefficient alphas) for all

    study variables. As can be seen, there was significant convergence between employeesand co-workers in their ratings of both CWB (r= .29, p < .01) and OCB (r= .29, p