Oxford Handbooks Online

Transcription

1 Oxford Handbooks Online Within-Person Approahes to the Study of Organizational Citizenship Behavior: Anteedents, Brent A. Sott, Fadel K. Matta, and Joel Koopman The Oxford Handbook of Organizational Citizenship Behavior Edited by Philip Podsakoff, Sott B. Makenzie, and Nathan P. Podsakoff Subjet: Psyhology, Organizational Psyhology Online Publiation Date: Sep 2016 DOI: /oxfordhb/ Abstrat and Keywords This hapter provides a review of the nasent (but growing) literature on organizational itizenship behavior (OCB) at the within-person level of analysis. We organize our review of the existing literature hronologially, disussing anteedents and onsequenes of within-person flutuations in OCB. After providing a narrative review of the literature, we provide a quantitative summary of the literature via meta-analysis, summarizing the within-person relationships between OCB and its most ommon within-person orrelates (i.e., positive affet, negative affet, job satisfation, stressors, strain, and task performane). Looking to the future of OCB at the within-person level of analysis, we suggest that researhers an ontribute to the domain by tailoring the measurement of OCB to the within-person level of analysis, better illuminating the ausal diretion between OCB and affet, larifying the relationship between OCB and ounterprodutive work behavior at the within-person level, expanding the dark side of within-person OCB, exploring between-person differenes in within-person OCB variability, and inorporating new theories. Keywords: organizational itizenship behavior, meta-analysis, within-person, experiene sampling, positive affet, negative affet, job satisfation, stressors, strain, performane Sine its introdution to the organizational behavior literature, the vast majority of researh on organizational itizenship behavior (OCB; Organ, 1988; Smith, Organ, & Near, 1983) and related onstruts, inluding prosoial behavior (George, 1991), organizational spontaneity (George & Brief, 1992), extrarole behavior (Van Dyne & LePine, 1998), proative behavior (Crant, 2000; Morrison & Phelps, 1999), and ontextual performane (Motowidlo & Van Sotter, 1994), has foused on identifying why some employees engage in higher levels of OCB than others (e.g., Organ & Ryan, 1995) and the onsequenes of their behavioral tendenies for outomes at both the individual and Page 1 of 34

2 organizational levels (e.g., Podsakoff, Whiting, Podsakoff, & Blume, 2009). Though researh at this between-person level of analysis has been ritial for understanding OCB, even the best employees sometimes do not engage in OCB; similarly, even the worst employees sometimes do engage in OCB. In other words, there are substantial and systemati flutuations in employees episodi, or momentary levels, of OCB. This is an important departure from traditional ross-setional researh, whih typially would view suh flutuations as transient error. The purpose of this hapter is to examine the emerging literature on OCB onduted at the within-person level of analysis. We first review the researh that has been onduted to date at this level of analysis, fousing on the anteedents and onsequenes of momentary engagement in OCB. We then provide a quantitative summary of this literature via meta-analysis in order to provide a snapshot of what we know as well as to lay the groundwork for where we might go to move this emerging stream of researh forward. In so doing, we draw omparisons to the broader stream of researh examining between-person differenes in OCB. Finally, and perhaps most important, we disuss the future of researh in this area. Issues that we takle inlude the measurement of OCB at the within-person level of analysis, the potential dark side of engaging in momentary OCB, between-person variability in OCB, and theoretial frameworks that an guide researh on OCB at the within-person level. Within-Person Researh on Organizational Citizenship Behavior The inorporation of experiene-sampling methodology (ESM; Larson & Csikszentmihalyi, 1983) has proven to be an invaluable tool for illuminating the dynami, within-person nature of OCB. Typially, ESM studies (sometimes referred to as diary studies) use systemati or random prompts to obtain observations of foal variables from individuals daily, or perhaps multiple times per day. Initially, ESM was used by researhers to apture moods and emotions (e.g., Alliger & Williams, 1993), whih by their nature are fleeting states and thus best examined in situ. It is perhaps not surprising then that the bulk of within-person researh on OCB has linked the onstrut to positive and negative affetive states (where affet is an umbrella term that enompasses both moods and emotions). Although sholars argued during the early 1990s that OCB and mood should be assoiated with eah other (e.g., George & Brief, 1992), the initial researh that followed using ESM aptured affetive states at the within-person level but assessed OCB Page 2 of 34

3 at the between-person level (e.g., Fisher, 2003), and therefore the relationship between affet and OCB was still assessed at the between-person level. This hanged with Sonnentag s (2003) publiation. In an ESM study of 147 employees from publi servie organizations who ompleted daily surveys for a 5-day period, she found that 46% of the variane in proative behavior (operationalized as personal initiative) was within-persons. Feelings of reovery in the morning were positively assoiated with reports of proative behavior at the end of the day, and work engagement mediated this relationship. Two years later, Miner, Glomb, and Hulin (2005) examined within-person OCB more diretly. Drawing from affetive events theory (AET; Weiss & Cropanzano, 1996), the authors argued that OCB at the momentary level represents what AET refers to as an affet-driven behavior (as opposed to a ognition-driven behavior, whih would be influened more by job attitudes). Their ESM study onsisted of 41 manufaturing employees who ompleted surveys eah day over a 2- to 3-week period. Following a morning survey that assessed initial mood, partiipants were signaled at four, stratified random times throughout the workday. Mood, or hedoni tone, was assessed with both positively (e.g., happy, pleased) and negatively valened (e.g., blue, sad) items. Selfreported OCB was assessed with two items that aptured both OCB direted toward the organization (OCBO) and OCB direted toward oworkers (OCBI) (e.g., Williams & Anderson, 1991), with respondents indiating in a yes/no format whether they were engaging in these behaviors at the time they were signaled. Partiipants ompleted 59% of the daily surveys, and 56% of the variane in hedoni tone was within persons (the perentage of within-person variane in OCB was not reported, though partiipants did report spending approximately 23% of their time helping oworkers). Although their predition of a relationship between hedoni tone and OCB was not supported, their investigation laid the groundwork for future researh treating OCB as an outome of affet. It is also noteworthy that Miner et al. (2005) ontrolled for hedoni tone in the morning, whih permitted an examination of whether a hange in hedoni tone was assoiated with OCB a point to whih we return later. Following the publiation of Miner et al. (2005), Ilies, Sott, and Judge (2006) onduted a 3-week ESM study with 62 partiipants from a variety of oupations. Toward the end of eah workday, partiipants ompleted measures of positive affet from the Positive and Negative Affet Shedule (PANAS; Watson, Clark, & Tellegen, 1988), job satisfation, and OCB, with 89% of all daily surveys ompleted. To assess OCB, the authors utilized 11 items from the widely used OCBI and OCBO sales developed by Lee and Allen (2002), eliminating items that would be less likely to vary daily (567). Twenty-nine perent of the variane in OCB was within persons, and both positive affet and job satisfation Page 3 of 34

4 were positively assoiated with daily engagement in OCB. Although the authors did not distinguish between OCBI and OCBO in their main analysis, a supplemental analysis revealed that OCBO was assoiated with both job satisfation and positive affet, while OCBI was assoiated with positive affet only, suggesting that OCBI may be more affet driven. Perhaps most interesting, in an analysis linking within- and between-levels of analysis, trait agreeableness (but not onsientiousness) moderated the within-person relationship between positive affet and OCB, suh that the relationship was more positive for those low in agreeableness. In ontrast, those high in agreeableness engaged in OCB regardless of their momentary mood. In all, the Ilies et al. (2006) study is important beause it reveals the omplex ways in whih stable, individual traits an interat with experiened states to predit momentary OCB (see also Mishel & Shoda, 1995). In a rather unique study of 44 flight attendants who ompleted surveys over three onseutive trips, with three surveys per trip, Xanthopoulou, Heuven, Demerouti, Baker, and Shaufeli (2008) examined the effets of state self-effiay and olleague support on extra-role (as well as in-role) performane. Extra-role performane was assessed with two items that referened doing more work than was required and helping olleagues. Partiipants ompleted 63% of the diary surveys, and 60% of the variane in extra-role performane was within persons. The results of their model, whih was inspired by the job demands resoures model (Demerouti, Bakker, Nahreiner, & Shaufeli, 2001), showed that state self-effiay, but not olleague support, was positively assoiated with extra-role performane. Moreover, state work engagement mediated this effet. Overall, their study provides some insight into the motivational proesses underlying momentary engagement in OCB, as opposed to purely affetive proesses. Several papers examining OCB within persons were published in In what has proven to be an influential study on the relationship between OCB and ounterprodutive work behavior (CWB), Dalal, Lam, Weiss, Welh, and Hulin (2009) surveyed 48 employees from a software ompany (Study 1) and 67 employees from a variety of oupations (Study 2) over a 3-week period (i.e., 15 workdays). In Study 1, the employees ompleted four surveys eah day. OCB was measured with six items adapted from existing soures. Similar to Ilies et al. (2006), the authors omitted items that were likely to our rarely, if at all (Dalal et al., 2009, p. 1056). Similar to Miner et al. (2005), employees indiated whether they had engaged in eah behavior (as opposed to the extent to whih they had engaged in eah behavior), and affet pleasantness was assessed in a bipolar fashion suh that negatively valened items were reverse-oded. However, in ontrast to results of Miner et al. (2005), affet pleasantness (at the urrent time point, but not the previous time point) was positively assoiated with OCB. OCB and CWB were not related. In Study 2, the employees ompleted two surveys eah day. OCB was assessed more Page 4 of 34

5 omprehensively, with six items referening oworkers, six items referening the supervisor, and six items referening the organization. The perentage of variane in OCB within individuals was similar for eah referent (44.2%, 50%, and 52%, respetively). Importantly, positive and negative affet were kept separate in the analyses (as opposed to the bipolar operationalization utilized in Study 1). Positive affet was positively assoiated with all three forms of OCB, while negative affet was not. Moreover, neither positive nor negative affet at the previous time point was assoiated with OCB. Again, the relationship between OCB and CWB was generally weak. Overall, given that lagged preditors were ontrolled in the regression equation, the results of the studies by Dalal et al. reveal that an inrease in affet (speifially, positive affet) was assoiated with an inrease in OCB. Their results also suggest the intriguing notion that, even at the momentary, within-person level of analysis, OCB and CWB are independent, distint onstruts, whih parallels findings at the between-person level of analysis (Dalal, 2005). Binnewies, Sonnentag, and Mojza (2009) drew from onservation of resoures theory (Hobfoll, 1989) to explain the relationship between the state of being reovered in the morning and engagement in OCB during that day. Ninety-nine publi servie employees ompleted surveys on poket omputers in the morning, immediately after work, and in the evening for four onseutive workdays, with 91% of all surveys ompleted. OCBI was measured with five items adapted from existing sales, and 50% of the variane was within persons. The results revealed that the state of being reovered was positively assoiated with OCBI. Moreover, there was a ross-level interation between individuals stable levels of job ontrol and reovery, suh that the within-person relationship between the state of being reovered and OCB was positive for those with high job ontrol but not signifiant for those with low job ontrol. Overall, the Binnewies et al. study suggests that resoures are important for daily engagement in OCB, and job ontrol is needed in order for employees to apitalize on days when their resoure pool has been replenished. Integrating affetive events theory with the transational model of stress (Lazarus & Folkman, 1984), Rodell and Judge (2009) investigated the within-person relationships among hallenge and hindrane stressors, emotions, and both OCB and CWB. Challenge stressors are job demands that reate potential for personal growth, whereas hindrane stressors are job demands that present obstales to personal growth (Cavanaugh, Boswell, Roehling, & Boudreau, 2000). One hundred partiipants from a variety of organizations were given 3 weeks to omplete 10 daily surveys toward the end of their workday. OCB was assessed with 11 items from Lee and Allen s (2002) measure, and 38% of the variane in OCB was within persons. Challenge stressors had a positive, indiret relationship with OCB through the affetive state of attentiveness. In addition, both Page 5 of 34

6 hallenge and hindrane stressors had a negative, indiret relationship with OCB through the affetive state of anxiety. No differenes were observed when OCB was broken down into OCBI and OCBO. As an aside, the authors also assessed CWB, and the withinindividual orrelation between OCB and CWB was negative and signifiant. The upshot of the Rodell and Judge (2009) study is that not all stressors are equal when it omes to daily engagement in OCB. Although hindrane stressors appear to be unequivoally bad, the relationship between hallenge stressors and OCB is more omplex, with offsetting effets on OCB via a relationship with an affetive state that is good for OCB (i.e., attentiveness) and an affetive state that is bad for OCB (i.e., anxiety). Examining anteedents of daily proative behavior, Fritz and Sonnentag (2009) had 172 ivil servie employees omplete morning and afternoon surveys for a 3-day period. Proative behavior was operationalized as taking harge (Morrison & Phelps, 1999). The perentage of variane in proative behavior that was within persons was not reported. Positive mood was assoiated with proative behavior on the same and the following workday. Interestingly, stressors in the form of time pressure and situational onstraints were also positively assoiated with proative behavior, suggesting that individuals took harge in an effort to alleviate their undesirable situation. Last but not least in 2009, Conway, Rogelberg, and Pitts had 80 partiipants from a variety of organizations use palm omputers to omplete five surveys eah day for 5 workdays, with 87% of surveys ompleted. Helping (OCBI) was assessed with a single item ( sine the last signal, did you voluntarily help someone else [in a way that was not an assigned duty]? ) (p. 328). In ontrast to the previously reviewed studies, only 13% of the variane in OCBI was within persons. Positive affet (assessed with four items) in the previous time period was not assoiated with OCBI. However, trait altruism (but not trait empathy) moderated the within-person relationship between positive affet and OCBI, suh that individuals high in altruism engaged in OCBI regardless of their positive affet, while individuals low in altruism engaged in OCBI as their positive affet inreased. This result is similar to the finding of Ilies et al. (2006) that agreeableness (of whih altruism is a faet) moderated the within-person relationship between positive affet and OCBI. In a supplemental analysis with positive affet as an outome instead of a preditor, OCBI was not assoiated with a hange in positive affet. Overall, the lak of signifiant withinperson results may have been due, in part, to the restrited amount of within-person variane in OCBI. Departing methodologially from prior researh, Binnewies, Sonnentag, and Mojza (2010) examined weekly relationships involving OCB and reovery. Similar to daily designs, 47.5% of the variane in OCB was within persons. Results indiated that the state of being reovered (assessed on Monday) was positively assoiated with engagement in Page 6 of 34

7 OCB that week, suggesting that OCB is more likely for individuals who take time to replenish their resoures over the weekend. Ohly and Fritz (2010) had engineers from an automotive manufaturer rate their daily engagement in proative behavior (operationalized as personal initiative) for a 3-day period. A total of 43.1% of the variane in proative behavior was within persons. Similar to the findings of Fritz and Sonnentag (2009), the authors found that hallenging harateristis of work (i.e., time pressure and job ontrol) were positively assoiated with proative behavior, and these relationships were partially mediated by hallenge appraisals. Miner and Glomb (2010) had 67 all-enter employees omplete one morning and four or five randomly sheduled surveys throughout the workday for 3 weeks, with 84% of surveys ompleted. OCB was measured with two items that aptured behavior direted toward the organization (OCBO) and behavior direted toward others (i.e., helping; OCBI). The amount of within-person variane in OCB was not reported. Affet ( hedoni tone ) was assessed in the morning and during the workday with both positively and negatively valened items, with negatively valened items reverse-oded (f. Miner et al., 2005). Engaging in OCB was not assoiated with a hange in hedoni tone, and metamood, whih aptures a person s general tendeny to attend to and regulate his or her emotions (Salovey, Mayer, Goldman, Turvey, & Palfai, 1995) did not moderate this within-person relationship. Although the authors found signifiant results with other preditors (e.g., work withdrawal), the lak of findings for OCB ould have been due to the measurement of affet, whih utilized a bipolar operationalization as opposed to keeping positive and negative affet separate (f. Dalal et al., 2009). Glomb, Bhave, Miner, and Wall (2011) did separate positive and negative affet in their ESM study of 68 managerial and professional employees. In an intriguing departure from prior work, Glomb et al. proposed that individuals engage in OCB in order to repair a negative mood, thereby positioning affet as an outome rather than an anteedent of OCB. The employees in their sample ompleted randomly signaled surveys (a morning survey used to apture baseline affet and four surveys throughout the workday) on palmtop omputers for 3 weeks. Approximately 53% of the surveys were ompleted. The altruism and ourtesy dimensions of OCB were assessed with multiple items, and similar to previous researh, the authors eliminated items less likely to our on a daily basis. Seventy perent and 87% of the variane in altruism and ourtesy, respetively, was within persons. The results of their ESM study showed that negative affet at the previous time period was assoiated with an inrease in altruism but not ourtesy. Positive affet at the previous time period was not assoiated with a hange in either form of OCB. Subsequently, engaging in both forms of OCB was assoiated with an inrease in positive affet. However, ourtesy was also assoiated with an inrease in Page 7 of 34

8 negative affet. Finally, there was a ross-level interation with extraversion suh that extraverts positive affet inreased more following their altruisti behavior. What is interesting and important about the Glomb et al. (2011) study is that it departs from the fous on OCB as solely an outome in within-individual studies. That is, not only might individuals engage in OCB in an attempt to repair their negative mood but also engaging in OCB may have downstream onsequenes. In two ESM studies, Halbesleben and Wheeler (2011) drew from onservation of resoures theory to understand the relationship between feelings of exhaustion and OCB. In Study 1, 199 partiipants from a variety of organizations took part in a weeklong study, ompleting an online survey eah day for five workdays. OCBI was assessed using Williams and Anderson s (1991) sale, and 31% of the variane in OCBI was within persons. Counterintuitively, exhaustion was positively assoiated with OCBI. This assoiation was more positive in relationships where the foal employee reeived more than she or he ontributed (i.e., in relationships haraterized by positive reiproity). In Study 2, 155 partiipants from a variety of organizations took part in a weeklong ESM study, similar in design to Study 1. Both OCBI and OCBO were assessed using the Williams and Anderson (1991) sale, and 35% of the variane in OCBI, and 22% of the variane in OCBO, was within persons. Again, exhaustion was positively assoiated with OCBI, and this relationship was stronger in relationships haraterized by positive reiproity. However, exhaustion was negatively assoiated with OCBO, and this relationship was also stronger in relationships haraterized by positive reiproity. The authors explained the rather ounterintuitive finding between exhaustion and OCBI by suggesting that when experiening a state of exhaustion, employees hoose more wisely where to invest their limited resoures, and OCBI is more lurative in this regard ompared to OCBO. Drawing from soial omparison theory (Festinger, 1954) and AET (Weiss & Cropanzano, 1996), Spene, Ferris, Brown, and Heller (2011) had 99 partiipants from a variety of oupations omplete an online survey eah day for a 2-week period, with 78% of surveys ompleted. OCB was measured with 14 items from the Lee and Allen (2002) OCBI and OCBO sales. Forty-three perent of the variane in OCB was within persons. Downward soial omparisons (where the foal employee pereives that she or he is better off than others) were assoiated with higher levels of OCB, but upward soial omparisons (where the foal employee pereives that she or he is worse off than others) were not assoiated with OCB. Both forms of soial omparisons had an indiret effet on OCB through positive affet. In addition, both of these within-person relationships were moderated by employees harateristi beliefs in a just world, suh that those who believe in a just world were more likely to engage in OCB when they made downward soial omparisons but less likely to engage in OCB when they made upward soial omparisons. Overall, the Page 8 of 34

9 Spene et al. (2011) study raises the interesting notion that how employees see themselves in relation to a target influenes whether they engage in OCB toward that target. Fay and Sonnentag (2012), using palmtop omputers, had 52 German employees from a variety of organizations partiipate in a week-long ESM study with three surveys per day. Employees ompleted 91% of the daily surveys, and 88% of the variane in OCB (operationalized by time spent on proative behavior) was within persons. Positive affet was assoiated with an inrease in time spent engaging in OCB. In ontrast, high levels of momentary experiened ompetene were assoiated with more time spent on ore task ativities and less time spent on OCB. The authors argued (but did not test diretly) that low feelings of ompetene motivate individuals to repair the negative state, and OCB is one way to aomplish this. Examining forms of support as preditors of OCB, Shreurs, Van Emmerik, Gunter, and Germeys (2012) had 56 Duth employees omplete a weekly survey for 3 weeks, ahieving a 95% response rate for the weekly surveys. OCB (extra-role performane) was assessed with seven items developed by Goodman and Svyantek (1999), and 60% of the variane in OCB was within persons. In terms of preditors, although supervisor support was positively assoiated with OCB, neither olleague support nor job seurity was assoiated with OCB. Wu and Parker (2012) had 58 undergraduates report on their proative behavior (measured by adapting a sale of proative personality [Bateman & Crant, 1993) on a monthly basis for 4 months. A total of 40.8% of the variane in proative behavior was within persons. The states of uriosity, ore self-evaluations, and future orientation were all assoiated with higher levels of proative behavior at that time. In addition, there were ross-level interations, suh that the within-person relationship between ore selfevaluations and proative behavior was weaker, while the within-person relationship between future orientation and proative behavior was stronger, for individuals with high levels of relationship anxiety. The authors onluded that people who worry about their soial relationships are hesitant to engage in proative behavior as a result of having a positive self-onept at that time. Barnes, Ghumman, and Sott (2013) examined the relationship between sleep quantity and daily engagement in OCB. Eighty-five employed university students were surveyed twie daily over the ourse of 5 workdays, and they ompleted 78% of the surveys. A preshift survey assessed sleep quantity, and a postshift survey assessed job satisfation and OCB over the ourse of the workday. OCB was measured using the 16-item Lee and Allen (2002) measure, and 30% of the variane in OCBI, and 20% of the variane in Page 9 of 34

10 OCBO, was within persons. Sleep quantity was positively assoiated with engagement in both OCBI and OCBO, and job satisfation mediated these relationships. Three studies on OCB at the within-person level were published in 2014, and they extend the aforementioned researh in several ways. First, in a unique fous on managers and their engagement in justie-relevant ations, Johnson, Lanaj, and Barnes (2014) drew from theory on regulatory resoures (e.g., Baumeister, Bratslavsky, Muraven, & Tie, 1998) to understand the onsequenes of engaging in proedural and interpersonal justie for managers themselves. Seventy-nine managers were surveyed in the morning and afternoon for 10 workdays (2 weeks), and they ompleted 69% of the daily surveys. Managers engagement in OCBI (helping others) was measured with four items from Fox, Spetor, Goh, Bruursema, and Kessler (2012), and 34% of variane in OCBI was within persons. Engaging in proedural justie was assoiated with an inrease in feeling depleted the next day, whereas engaging in interpersonal justie was assoiated with a derease in feeling depleted the next day. Moreover, depletion mediated the effets of justie behaviors on OCBI. In a supplemental analysis demonstrating support for the authors proposed ausal diretion, OCBI was not assoiated with either justie or depletion the next day. Finally, the positive, indiret effet of interpersonal justie on OCB via depletion was stronger for managers low in extraversion and for managers high in neurotiism. The findings for depletion, in whih feelings of depletion were assoiated with lower levels of OCB, are somewhat in ontrast to the findings of Halbesleben and Wheeler (2011), who found that exhaustion was positively assoiated with engagement in OCBI. In an intriguing analysis of within-person variation in personality traits, whih historially have been thought to be stable, Judge, Simon, Hurst, and Kelley (2014) olleted data from 122 partiipants from a variety of oupations who took part in a 2-week ESM study. Partiipants ompleted 86% of the daily surveys. OCB was measured with 12 items tapping both OCBI and OCBO from the Lee and Allen (2002) sale, and 48% of the variane in OCB was within persons. Engagement in OCB was positively assoiated with next-day extraversion, agreeableness, and openness to experiene. Thus, the more individuals engaged in OCB, the more they reported behaving in an extraverted, agreeable, and open manner the following day. In ontrast, the effets of personality states on next-day OCB were not signifiant. These results suggest that the approahoriented nature of OCB triggers orresponding, approah-oriented personality states in individuals. As suh, they are yet another example of how within-person researh an differ dramatially from between-person researh, whih historially has treated personality as a stable anteedent of OCB (e.g., Organ & Ryan, 1995). Page 10 of 34

11 Departing from previous researh that has foused on positive and negative affet as broad states, Spene, Brown, Keeping, and Lian (2014) examined the disrete emotion of state gratitude and its relationship with OCB at the within-person level. The authors onduted two ESM studies. The first onsisted of 67 partiipants from various oupations who ompleted a morning and afternoon survey eah day for 5 workdays, with 80% of the surveys ompleted. OCB was assessed with 14 items from the Lee and Allen (2002) sale, and 38% of the variane in OCB was within persons. State gratitude was positively assoiated with engagement in OCB. In Study 2, 104 partiipants from an array of oupations ompleted an afternoon survey eah day for 10 workdays, with 63% of the surveys ompleted. Dalal et al. s (2009) 18-item sale was used to assess OCB. In terms of variane, 33%, 41%, and 34% of the variane in OCB direted toward the supervisor, oworkers, and organization, respetively, was within persons. Over and above positive affet, state gratitude was positively assoiated with OCB direted toward one s supervisor and oworkers, but not toward one s organization. Overall, their study demonstrates the utility of a disrete emotion approah, and it is one of the few studies that have taken suh an approah. In an interesting analysis of the effets of pain on disretionary behavior, Christian, Eisenkraft, and Kapadia (2015) onduted two ESM studies examining promotive, extrarole behaviors within persons. In the first, 85 employees experiening hroni pain took part in a 3-week study whereby they ompleted morning and afternoon measures. Partiipants ompleted approximately 78% of those surveys, and 49% of the variane in extra-role behavior was within persons. Daily pain was indiretly assoiated with less extra-role behavior via lower work engagement, but there was not a signifiant, indiret effet through resoure depletion. In the seond study, the authors investigated intentions to engage in extra-role behaviors (66% of the variane was within persons). Results of that study, whih utilized arhival data from 650 partiipants who provided 6,820 observations over the ourse of a week, were similar to the results of the first study, exept that the indiret effet of pain on extra-role behavior through resoure depletion was signifiant. Overall, their studies demonstrate the important of human energy for momentary engagement in OCB, even over and above moods and emotions. Gordon, Demerouti, Bipp, and Le Blan (2015) also examined work engagement in onert with OCB (ontextual performane). Their 5-day ESM study of 49 nurses ahieved a 60% daily response rate, and 60% of the variane in OCB was within persons. Intuitive deision making was positively assoiated with OCB, partiularly for nurses with high levels of work engagement. Sonnentag and Starzyk (2015), in a 2-week ESM study of 153 employees who ompleted two surveys per day, examined a dual, affetive pathway model to engagement in Page 11 of 34

12 proative behavior (operationalized as issue identifiation and issue implementation). Roughly half of the variane in these forms of proative behavior was within persons. In support of their dual-pathway approah, pereptions of prosoial impat were assoiated with an inrease in positive affet later that day, whih in turn was assoiated with greater issue implementation (but not identifiation). In ontrast, pereptions of situational onstraints were positively assoiated with an inrease in negative affet later that day, whih in turn was assoiated with greater issue identifiation (but not implementation). Overall, their findings suggest that different aspets of OCB may be triggered by both positive and negative affet. Moreover, their findings appear to be onsistent with researh on emotion showing that negative affet auses people to reflet and plan (e.g., Izard & Akerman, 2000), while positive affet auses people to engage and exeute (e.g., Fredrikson, 2001). In an illustration of the infany of researh on within-individual OCB, Trougakos, Beal, Cheng, Hideg, and Zweig (2015) proposed a model that ontradits the previously disussed manusript by Halbesleben and Wheeler (2011). Drawing on ego depletion theory, Trougakos et al. (2015) hypothesized that daily surfae ating inreases emotional exhaustion and dereases engagement in OCBI. Using three items from Williams and Anderson s (1991) OCBI subsale (the perentage of variane that was within individuals was not reported), these authors found support for their model. Note that the diretion of the relationship proposed between emotional exhaustion and OCBI is opposite to that proposed and found by Halbesleben and Wheeler (2011). This ontradition is interesting due to several notable differenes between the studies. First, although both used items from Williams and Anderson (1991), Trougakos and olleagues used only three items, whereas Halbesleben and Wheeler (2011) used all seven. Could there be differenes in the omitted items that influene the relationship that substantially? A more likely ulprit is that OCBI was measured from different soures between the studies. Speifially, Halbesleben and Wheeler (2011) relied upon self-reports, as is ommon in withinindividual studies of OCB. Trougakos and olleagues instead obtained a daily report from one of the foal employee s oworkers. We return to some of the impliations of this deision in a subsequent setion. Finally, two reent studies illustrate the evolution of ESM researh on OCB sparked by Glomb et al. (2011) by fousing not on anteedents, but instead on downstream outomes. In a follow-up to their 2011 study desribed earlier, Halbesleben and Wheeler (2015) investigated a mehanism by whih OCBI may serve as a resoure investment in others. These authors adopted a dyadi ESM method in whih 177 pairs of employees ompleted daily surveys for a 1-week period. Instead of the Lee and Allen (2002) measure, OCBI was assessed using the seven-item OCBI subsale from Williams and Page 12 of 34

13 Anderson (1991) (the perentage of OCB variane that was within persons was not reported). In support of their hypotheses, Halbesleben and Wheeler showed that on days when a given employee engaged in higher levels of OCBI, his or her oworker pereived higher levels of resoures available through that relationship (operationalized as soial support). This higher level of support was assoiated with higher levels of trust that day and inreased reiproation of OCBI the following day. Finally, in support of the notion of resoure gain spirals proposed by onservation of resoures theory (Halbesleben, Neveu, Paustian-Underdahl, & Westman, 2014; Hobfoll, 1989), these authors showed that levels of support, trust, and OCBI inreased over the ourse of the study. Koopman, Lanaj, and Sott (2016) similarly foused on the outomes of OCB to ators, but these authors departed from extant within-individual researh on OCB in an important way. In the aforementioned studies, OCB has generally been oneptualized as a positive behavior (e.g., assoiated with positive affet, a means of investing resoures, a mehanism for mood repair). However, OCB researh at the between-individual level of analysis has identified a potential dark side to these behaviors (for a review, see Bolino, Klotz, Turnley, & Harvey, 2013). Koopman et al. (2016) developed a similar theory at the within-individual level and tested a model illustrating the bright and dark sides of daily engagement in OCB. In their ESM study, 82 employees from a variety of oupations ompleted three online surveys eah day for a period of 2 weeks (10 workdays), with a 91% day-level response rate. OCBI was assessed with six items reported in Dalal et al. (2009), and 54% of its variane was within persons. On the bright side, OCBI was assoiated with an inrease in positive affet, and positive affet mediated the relationship between OCBI and both affetive ommitment and job satisfation, but not emotional exhaustion. On the dark side, OCBI was assoiated with feelings of redued goal progress, and goal progress mediated the relationship between OCBI and the three well-being outomes (i.e., affetive ommitment, job satisfation, and emotional exhaustion). Interestingly, these opposing relationships were moderated by individuals trait regulatory fous (Higgins, 1997), suh that the positive relationship between OCBI and positive affet was stronger for those with a high promotion fous, while the negative relationship between OCBI and pereptions of work goal progress was stronger for those with a high prevention fous. Combining these analyses, evidene for moderated mediation of the first stage of the indiret effet was found. Overall, the results of the Koopman et al. study suggest that momentary engagement in OCB has both benefits and drawbaks. Although it boosts positive affet, it also interferes with goal progress on ore work tasks beause engagement in OCB must neessarily ome at the expense of task performane at a given moment in time. Thus, when it omes to answering the question as to whether engaging in OCB is good or bad for a given employee, it appears that the question should be hanged to for whom is OCB good or bad, and why (Koopman et al., 2016, p. 28) a point to whih we return later. Page 13 of 34

14 A Mini Meta-Analysis on Within-Person Organizational Citizenship Behavior Researh Having desribed the existing researh on within-person anteedents and onsequenes of OCB, we wanted, to the extent possible, to provide a quantitative summary of that researh. Therefore, we meta-analyzed those within-person relationships involving OCB that have been examined most frequently. Before we get to those relationships, we wanted to get an idea of the amount of variane in OCB that is within persons. As reviewed earlier, the perentage of variane in OCB that is within persons varies quite a bit from study to study. In fat, the range was approximately 25% to 88%. The average amount of within-person variane was 45.5% (SD =.15). For OCBI, it was slightly higher, at 48.2% (SD = 15.7%). For OCBO, it was slightly lower, at 36.0% (SD = 12.0%). In onsidering these differenes, it ould be the ase that OCBI varies more within persons beause ats suh as helping others may depend on interation opportunities with speifi oworkers (or with one s supervisor). Similarly, within-person variane in OCBI may be higher beause of greater flutuations in oworkers (or one s supervisor s) need for help on a given day. In ontrast, OCBO may vary more between persons beause one s relationship with the organization is more stable over short periods of time, thus refleting a general tendeny to be a good itizen. Caution, however, should be taken with respet to the differenes between OCBI and OCBO beause the results for OCBO were based on only six studies. How do the variane results for OCB ompare to within-person variane in other onstruts that have been frequently studied at the within-person level? Quite well, as it turns out. In the studies on OCB that we reviewed earlier and inluded in our metaanalysis, state-positive (M = 51.4%, SD = 16.5%) and state-negative (M = 57.2%, SD = 10.1%) affet exhibited similar amounts of within-person variane. Beause one would expet affetive states, whih are often fleeting and ephemeral, to exhibit a substantial amount of within-person variane, the similar amount of within-person variane for OCB is noteworthy. For the sake of further omparison, the within-person variane in other onstruts inluded in our review was as follows: job satisfation (M = 29.4%, SD =.05), task performane (M = 43.8%, SD = 20.4%), stressors (M = 43.0%, SD = 14.1%), and strain (M = 61.5%, SD = 17.7%). Overall, these results demonstrate the importane of examining OCB within persons over time. After all, if sholars only onentrated on between-person variane in OCB, whih was the ase only a deade ago, they would be missing half of the story. Page 14 of 34

15 Turning now to within-individual relationships involving OCB, we onduted our metaanalysis following Hunter and Shmidt s (2004) guidelines. Our results inlude the sample-size-weighted point estimate of the study orrelations (r), a 95% onfidene interval around that point estimate, the number of studies (k), and the umulative sample size (N). In addition, we report the orrelation after orreting for unreliability in both the preditor and the riterion (r ), with orretions performed at the independent sample level (i.e., prior to meta-analyzing the orrelations). For studies that did not inlude reliability information, we substituted the weighted average from the studies that did report that information. Finally, we report both the standard deviation of the orreted meta-analyti orrelation (SDr ), the 80% redibility interval (whih provides an estimate of the variability in individual orrelations aross the studies), and the perentage of variane in the orrelations explained by artifats. Together, these latter three estimates an be used to suggest the presene of moderators. Table 1 presents the results of our mini meta-analysis ( mini simply beause there are not a large number of studies on OCB at the within-person level, though that number is inreasing). Unfortunately, there were not enough studies that measured OCBO to breakdown our meta-analyti results by referent (i.e., OCBO versus OCBI). For the sake of omparison, in disussing our results, we also present orrelations from prior metaanalyses at the between-person level. As shown in Table 1, OCB is positively orrelated with positive affet (r =.18; r =.22), and the 95% onfidene interval for this relationship exluded zero. OCB is also positively (albeit weakly) orrelated with negative affet (r =.07; r =.08), as the 95% onfidene interval exluded zero. Shokley, Ispas, Rossi, and Levine (2012) reported a between-person orrelation of r =.32 for statepositive affet and r =.02 for state-negative affet. Interestingly, the positive orrelation between OCB and negative affet supports the view that OCB is used on a momentary basis as a form of mood repair (Glomb et al., 2011). The stronger results for positive affet ompared to negative affet also support the view that positive and negative affet should be separated when examining their relationships with OCB (as opposed to ombining positive and negative affet items into a single hedoni tone measure). Page 15 of 34

16 Table 1 Meta-Analyti Results of Within-Person Correlates of Organizational Citizenship Behavior Organizational Citizenship Behavior Correlate r (95% CI) k (N) r (80% CV) SDr (%V ) art Positive affet.18 (.11,.24) 9 (7,476).22 (.04,.39).14 (8.86) Negative affet.07 (.01,.12) 6 (5,606).08 (.01,.17).08 (26.21) Job satisfation.27 (.14,.41) 3 (2,154).33 (.15,.50).15 (8.08) Stressors.24 (.12,.35) 6 (3,003).32 (.10,.53).18 (10.23) Strain.02 (.15,.19) 5 (3,650).02 (.27,.30).23 (3.98) Task performane.23 (.11,.35) 5 (2,432).32 (.12,.52).17 (12.93) CI, onfidene interval around unorreted population orrelation; CV, redibility interval around weighted orreted mean orrelation; k, number of studies; N, umulative sample size; %Vart, perentage of variane in r explained by study artifats; r, unorreted population orrelation; r, orreted population orrelation. OCB is also positively orrelated with job satisfation (r =.27; r =.33). Putting these results together with the results for positive affet supports the notion from affetive events theory that OCB, if onsidered an outome of affetive events, is likely both an affet-driven behavior and a ognition-driven behavior (see Weiss & Cropanzano, 1996). Page 16 of 34

17 Similarly, LePine, Erez, and Johnson (2002) reported a between-person orrelation of r =.24. Table 1 also shows that OCB is positively assoiated with stressors (r =.24; r =.32), as the 95% onfidene interval for this orrelation does not inlude zero. Here, stressors represent a variety of fators, inluding workload, job inseurity, and job demands, that both hallenge and hinder. Between-person researh, foused on stressors suh as role ambiguity, role onflit, and role overload, have reported orreted orrelations ranging from r =.06 to r =.15 (Eatough, Chang, Miloslavi, & Johnson, 2011). With the exeption of hallenge stressors, between-person researh has generally deemed the stressors inluded in the within-person researh we reviewed to be bad for employee well-being. On the one hand, the positive, within-person relationship between OCB and stressors suggests that on days when individuals are faed with a greater than usual level of stressors (and hene feel stressed), they invest their remaining resoures into OCB, perhaps with the hope that suh behaviors will be reiproated (f. Halbesleben & Wheeler, 2011). On the other hand, OCB is not signifiantly assoiated with strain (r =. 02; r =.02). This differs from between-person meta-analyti estimates, whih have reported a negative orrelation (r =.16) between OCB and strain (speifially, emotional strain) (Chang, Johnson, & Yang, 2007). It should be noted that the relationship between OCB and strain has the lowest perentage of variane explained by study artifats (at 3.98%), whih suggests the presene of moderators. In all, future researh is needed to better understand how momentary levels of stressors, stress, and strain relate to engagement in OCB. Finally, OCB is positively orrelated with task performane (r =.23; r =.32), whih, similar to findings at the between-person level of analysis (r =.37 for OCB; r =.39 for OCBO; Podsakoff et al., 2009), supports the view that these two dimensions of job performane are distint. Issues and Future Diretions for Within-Person Researh on Organizational Citizenship Behavior Having reviewed what we know about OCB at the within-person level of analysis, we now disuss issues in the study of OCB at this level of analysis, as well as potential diretions for future researh in this area, in order to lay the groundwork for where we might go. Page 17 of 34

18 Measurement of Organizational Citizenship Behavior Within Persons Over Time The first issue deals with the measurement of OCB at the within-person level of analysis, and there are several subissues within this larger issue. First, as noted in our review, the majority of within-person investigations of OCB have relied on the OCBI and OCBO sales developed by Lee and Allen (2002). Sometimes OCBI and OCBO are kept separate, and other times they are ombined into an overall OCB sale. Although we do not take exeption with examining overall OCB, the potentially bigger issue in our mind is that studies differ in the items that they take from these sales. To be fair, most of the withinperson studies to date have used the majority of items from the Lee and Allen (2002) sales, but when items are eliminated, it is diffiult to make omparisons aross studies beause the full sale is not provided. Thus, we enourage researhers using existing sales to be lear about whih items are inluded and whih items are exluded. Alternatively, future researh may benefit from using the sales reported in Dalal et al. (2009), who developed sales of OCB suitable for within-person researh by drawing from different soures and provided the items in an Appendix. The use of these sales, whih were tailored for a within-person fous, ould better allow for apples to apples omparisons aross studies. That said, a potential downside is that the entire measure onsists of 18 items (with OCB toward oworkers, the supervisor, and the organization), and therefore the development of a shorter, overall OCB sale would be welomed given the pratial limitations of ESM. Seond, when items are exluded from existing sales suh as Lee and Allen (2002), authors have done so based on the notion that some forms of OCB are unlikely to vary on a daily basis. Suh exlusions are reasonable given that these sales were developed to apture differenes in OCB between persons. For example, it probably is unlikely that one would observe variation over a short period of time for a behavior suh as adjust your work shedule to aommodate other employees requests for time off or attend funtions that are not required but that help the organizational image (Lee & Allen, 2002). That being said, this is ultimately an empirial question that researh has yet (but needs) to answer. Moreover, one items are eliminated, the question is whether this reates defiieny in the measure of OCB. One argument is that it does reate defiieny beause the full sale is not being used, yet those items are representative of the onstrut of OCB. However, an alternative argument is that the elimination of items that vary little over short periods of time does not reate defiieny beause the onstrut of OCB differs slightly aross level of analysis. In other words, over short periods of time, suh as hours or days, some of the behaviors thought to represent OCB are so rare as to be virtually irrelevant. Thus, it may be that OCB is not ompletely isomorphi aross Page 18 of 34

19 levels of analysis, with the degree of isomorphism between within-person OCB and between-person OCB inreasing as the period of time in between measurements of within-person OCB inreases. We noted earlier that shorter sales may be neessary at times given the demanding nature of ESM studies. To get around this issue, some researhers (e.g., Dalal et al., 2009) keep a ommon ore of items on eah momentary survey yet yle other items in. On the one hand, there are several benefits to this tehnique: a larger aspet of the onstrut spae is aptured, reduing potential defiieny that would be reated if items were eliminated, and it may redue priming effets beause partiipants are faed with a slightly different set of items eah time (Hulin & Judge, 2003; Miner et al., 2005). On the other hand, the slight differenes aross sales make it diffiult to know whether a relationship at a given moment in time (e.g., an employee reports high levels of both positive affet and OCB on Monday; on Tuesday that same employees reports high positive affet but low OCB) is being aurately aptured or is due to differenes in item ontent (the speifi types of OCB that this employee engages in was on Monday s sale but not Tuesday s). A final issue with regard to the measurement of OCB deals with self-report. In betweenindividual researh, it is not overly diffiult to have independent raters (e.g., supervisors or oworkers) assess the OCB of a foal employee. Indeed, the inlusion of self-reported OCB in between-persons researh is sometimes viewed as a major limitation. However, in within-persons researh, partiularly in intense ESM studies over short periods of time, independent assessments are unrealisti and infeasible to obtain. Moreover, it may be the ase that foal employees are the best soures for information about momentary OCB beause independent raters suh as supervisors may have limited observational opportunities. Fortunately, some auses of same-soure bias an be ruled out by withinperson designs, by individually mean-entering the data, whih removes auses of samesoure bias suh response tendenies and trait affet (Podsakoff, MaKenzie, Lee, & Podsakoff, 2003). Nevertheless, within-person investigations of OCB would benefit by inorporating other data besides self-report. Unobtrusive measures may be one promising area, partiularly in laboratory settings. An alternative to unobtrusive measures would be to seek daily ratings from oworkers as we disussed earlier. Though diffiult in an ESM study, suh designs are not impossible (e.g., Halbesleben & Wheeler, 2015; Trougakos et al., 2015). Moreover, far from being simply a methodologial point, suh investigations may be neessary to establish the theoretial domain of the daily OCB onstrut. As noted earlier, OCB may differ somewhat at the within-individual level (ompared to the between-individual level). Given the ontradition in the Halbesleben and Wheeler (2015) and Trougakos et al. (2015) papers disussed earlier, perhaps measurement from different soures exaerbates suh Page 19 of 34

20 differenes further. That is, at the aggregate, between-individual level perhaps employees an spread their OCB around so that the partiular oworker providing a measurement is unimportant. Contrast this with the daily level where an employee may engage in a high level of OCB (and would thus self-report this), but the partiular oworker providing the measurement was not the reipient of that OCB and therefore reports that the foal employee engaged in little OCB. Future researh is neessary to determine how suh situations might influene within-person OCB relationships and whether other reports of OCB are even appropriate in ESM studies. Causal Diretion of Affet Organizational Citizenship Behavior Relationships Muh of the researh on affetive states and OCB was prediated on affetive events theory (Weiss & Cropanzano, 1996). AET positions affetive reations to events as anteedents of behavior; thus, it is perhaps not surprising that studies using AET as a theoretial framework to understand within-person OCB have treated OCB as an outome of affet. For example, Dalal et al. (2009, p. 1053) noted that state affet is the primary anteedent of [OCB and CWB]. Spetor and Fox s (2002) model of voluntary work behavior also positions affet as an anteedent of OCB and CWB. Yet it is also reasonable (and reent studies have shown) that affet may be a onsequene of engaging in OCB (e.g., Glomb et al., 2011; Koopman et al., 2016). Thus, although OCB may indeed be a onsequene of affetive events, OCB may also onstitute an affetive event in and of itself that eliits emotion (most likely, positive affet). Given that ESM studies are now often olleting multiple surveys per day, lagged analyses and the examination of hange sores are possible, allowing researhers to better illuminate the ausal diretion between OCB and other onstruts. The Relationship Between Organizational Citizenship Behavior and Counterprodutive Work Behavior at the Within-Person Level Not inluded in our meta-analysis due to an insuffiient number of studies, the withinperson relationship between OCB and CWB is important to examine. Dalal (2005), in a meta-analysis at the between-person level of analysis, found that OCB and CWB were only moderately and negatively orrelated, and the two onstruts exhibit differential patterns of relationships with other variables. Thus, one might onlude at the betweenperson level of analysis that these two onepts are distint, suh that a person who engages in high levels of OCB may (or may not) engage in high levels of CWB. The independene of OCB and CWB was also the onlusion of Dalal et al. (2009), who found Page 20 of 34

21 that the two onstruts were unrelated at the within-person level. In ontrast, Rodell and Judge (2009) reported a negative (albeit small) orrelation between OCB and CWB at the within-person level. Although these findings appear to mirror researh at the betweenperson level, we wonder whether they are the result of being unable to apture engagement in OCB and CWB in the same moment in time. That is, by asking about engagement in OCB and CWB over the ourse of the day (Rodell & Judge, 2009) or sine the previous survey (Dalal et al., 2009) and thereby providing partiipants with a range of time, we allow for the possibility that an employee s OCB and CWB are unrelated. For example, over the ourse of a day, an employee would have time to engage in both OCB and CWB, high levels of one but not the other, or neither behavior at all. If, however, we dug down to the atual moment of behavioral exeution, would we still find OCB and CWB to be independent? We suspet not. Instead, we expet that OCB and CWB (and task performane, if one wants to apture the domain of job performane) to display strong, negative orrelations with eah other, perhaps to the point of being opposite ends of the same ontinuum. This is beause the more fine-grained the time frame beomes, the more one behavior must be performed at the expense of the other. Thus, it seems highly unlikely that both behaviors an be performed simultaneously. In sum, to provide a more definitive answer to the question as to whether a person who engages in OCB on a partiular oasion must also refrain from CWB on that oasion (Dalal et al., 2009, p. 1052), it may be that qualitative data or diret observation is needed to apture the preise behaviors that individuals are engaging in at a given point in time. Expanding the Dark Side of Within-Person Organizational Citizenship Behavior As disussed earlier, the Koopman et al. (2016) paper differed from the other manusripts that have investigated OCB at the within-individual level by drawing attention to the potential dark side of these behaviors. That OCB may have a dark side is already well known, however, theoretial and empirial investigations here have foused on issues suh as how OCB onflits with aggregate task performane or long-run areer trajetories (e.g., Bergeron, 2007; Bergeron, Shipp, Rosen, & Furst, 2013; Rubin, Dierdorff, & Bahrah, 2013). Koopman and olleagues illustrated that engaging in OCBs entails tradeoffs at the daily level, too. This highlights a number of potentially interesting opportunities for future researh. Glomb et al. (2011) noted that managers should view engaging in OCB as a means to promote well-being among employees. Is this reommendation ontradited by the findings of Koopman et al. (2016)? It would seem so, given that these authors linked OCB to redued well-being due to a lak of progress toward daily work goals. However, these Page 21 of 34

22 authors also noted an inrease in well-being through positive affet that was omparatively larger than the redution in well-being through a lak of work-goal progress, partiularly for promotion-foused employees. Should the take-home message then be that, in spite of some downsides, engaging in OCB has a net benefit for employees, partiularly in light of the potential soial benefits (e.g., Halbesleben & Wheeler, 2011, 2015)? This position has some merit; however, if employees onsistently fall short of their work goals, then in the long run they may experiene the areer-related onsequenes of OCB noted by others (e.g., Bergeron, 2007). Future researh that blends an individual s daily experienes of OCB with the longer run impliations of this behavior is needed in order to provide reommendations to both managers and employees. There is another aspet to the dark side of OCB that has been almost entirely overlooked in researh at both the within- and between-individual level: Who is on the reeiving end of this helping behavior? This shift in fous is important beause there are reasons to think that OCB ould have a double dark side that is, not only might it have negative impliations to the ator, but it might also be detrimental to the reeiver. For example, Deelstra and olleagues (2003) showed that reipients of helping behavior imposed upon them had negative reations in terms of both redued self-esteem as well as inreased heart rate. Thus, not only might helping behavior be bad for the ator, but it might also be bad for the reeiver. Of note here is that the helping behavior was imposed upon reipients, and so future researh should examine these effets more losely. One final opportunity for future researh on the dark side of OCB that we wish to disuss maintains the dual fous on ator and reeiver but also highlights potential negative onsequenes to the organization as well. Researh in other domains has identified that some employees may be onsidered ore or more tightly linked to the overall performane of the team (Humphrey, Morgeson, & Mannor, 2009, p. 49). Consider then several possible senarios. Perhaps a non-ore employee engages in OCB toward a ore employee on a given day. Although there may be a redution in performane to the ator (e.g., Barnes, Hollenbek, Wagner, DeRue, Nahrgang, & Shwind, 2008; Koopman et al., 2016), the overall ontribution to team or organizational performane might be a net positive if the help failitates the work of the ore employee that day. However, onsider the reverse situation, where a ore employee engages in OCB toward a non-ore employee. This situation might result in an overall net loss in produtivity on that day. Thus, future researh on OCB at the within-individual level ould attempt to identify who is doing the helping, and how this might impat the individuals involved in the behavior as well as the organization. Page 22 of 34

23 Using Experiene-Sampling Methodology to Study Between-Person Differenes in Within-Person Variability In addition to omplementing researh on between-individual differenes in average levels of OCB with the study of within-individual flutuations in OCB over time, researhers should also utilize ESM methods to merge these two streams of researh by exploring between-individual differenes in the stability of OCB over time. This type of approah has furthered our understanding of numerous organizational phenomenon inluding self-esteem (e.g., Kernis, Cornell, Sun, Berry, & Harlow, 1993), personality (Fleeson, 2001), interpersonal trust (Fleeson & Leiht, 2006), and emotional labor (Sott, Barnes, & Wagner, 2012). Variability in organizational phenomenon has been assoiated with both benefiial (e.g., Kernis et al., 1993) and detrimental outomes (e.g., Sott et al., 2012), and it is unlear whether OCB variability will be benefiial or detrimental for employees. On the one hand, OCB variability may be benefiial for an employee beause it may lessen the likelihood of job reep (whih ours when others expet employees to engage in OCB as an ongoing part of their role expetations; Van Dyne & Ellis, 2004) and esalating itizenship (whih ours when engaging in OCBs beomes so normative that employees feel obligated to ontinually inrease OCBs to be seen as going the extra mile; Bolino & Turnley, 2003). For example, when OCB is performed onsistently (as opposed to variably), it is more likely to be expeted and onsidered the norm. On the other hand, OCB variability may be detrimental for an employee if third parties (e.g., oworkers and supervisors) view these behaviors as impression management. For example, an employee who only performs OCB in the presene of his supervisor may not be viewed as favorably as an employee who performs OCB more onsistently and regardless of the situation. Overall, we see between-person differenes in within-person OCB variability as a partiularly fruitful area for future researh. Inorporating Other Theories to Understand Within-Person Organizational Citizenship Behavior To date, sholars investigating OCB at the within-individual level have tended to rely on affetive events theory or onservation of resoures theory as their overarhing theoretial frameworks. However, given the relative nasene of researh in this area, we feel that sholars may gain interesting insights into OCB at this level of analysis by drawing upon alternative theoretial frameworks. Thus, we onlude our hapter with a brief disussion of how several suh theories might be leveraged to inrease our understanding of daily OCB. Page 23 of 34

24 As reviewed earlier, Trougakos et al. (2015) reently applied ego depletion theory to their investigation of surfae ating as an anteedent of OCBI. One hallenge to withinindividual OCB researh is that onservation of resoures theory and ego depletion theory are often applied in very similar ways (e.g., that purposeful ats an redue available resoures). However, these theories are not idential. Conservation of resoures theory, for example, speifies that ertain events may be resoure generating as well as being resoure onsuming. Ego depletion theory does not learly artiulate this generation mehanism; however, it may possibly be more suited to theorizing at the momentary level of analysis where the depletion of available self-regulatory resoures might be the proximal ause of either ation or ination. Future researh ould therefore attempt to parse the mehanisms assoiated with these two theories to better understand how OCB influenes an individual s resoures. Moving beyond these theories, however, we think that there are opportunities to apply a number of other theoretial lenses to within-individual OCB to better understand both why employees engage in OCB, as well as the onsequenes of this behavior to ators and reipients. One option is moral liensing theory (Merritt, Effron, & Monin, 2010). Briefly, appliation of this theory suggests that a positive relationship would be found between OCB and CWB as a result of the aumulation or redution of moral redits. Klotz and Bolino (2013) reently applied this theory to OCB at the between-individual level of analysis, but would similar effets be found at the within-individual level as well? As our disussion of OCB and CWB earlier suggests, it is unlikely that a positive assoiation would be found in onurrent measurement of these onstruts. However, if a positive relationship were found over the ourse of a day or perhaps even from one day to the next, then perhaps a moral liensing framework ould be applied to explain this relationship. Building on the findings of Koopman et al. (2016) and a meta-analysis by Lanaj, Chang, and Johnson (2012), regulatory fous theory ould potentially be applied to understanding daily OCB. Lanaj et al. (2012) found that people with high trait promotion fous engage in more OCB, in general, and Koopman et al. (2016) found that high trait promotion fous amplified the positive affetive outomes of OCB. Combining these two findings, perhaps individuals primed with a high state promotion fous would be more likely to see OCB as a means of enhaning soial relationships (e.g., Halbesleben & Wheeler, 2015) and thus engage in more OCB ompared to individuals primed with a high state prevention fous, as suh individuals might be more attuned to the impliations of taking time away from assigned tasks to engage in OCB. Continuing the theme introdued earlier regarding looking at the reipients of OCB on a daily basis, researhers ould draw upon theory on self-esteem threat (Fisher, Nadler, & Page 24 of 34

25 Whither-Alagna, 1982) as one explanatory framework. Deelstra et al. (2003) drew upon this theory to explain why reipients of help reated negatively, and suh an explanation ould be similarly inorporated at the within-individual level. However, aording to Fisher et al. (1982), help need not always be interpreted as a threat; instead, helping an be seen as supportive as well (for example, see: Halbesleben & Wheeler, 2015). Thus, self-esteem threat ould be used to model both the bright and dark sides of reeiving help as well and explain when, why, and for whom reeiving help threatens one s selfesteem, and when it does not. Eah of the earlier disussions fouses on those who reeive help in the workplae. What about those who do not reeive help? Here, theory on soial omparisons (Wood, 1996) may be relevant. For example, do employees make soial omparisons with their oworkers about the amount of help they reeive? Researh on LMX soial omparisons suggests that employees do make soial omparisons with their workgroup peers about the resoures they reeive from others (e.g., Vidyarthi, Liden, Anand, Erdogan, & Ghosh, 2010). Moreover, soial omparison theory has already been a useful lens for understanding within-individual flutuations in OCB (e.g., Spene et al., 2011). As suh, researh ould explore the influene of daily soial omparisons about the amount of help employees reeive from others as well as when upward OCB soial omparisons result in ontrastive (aversive) or assimilative (omforting) soial omparison reations (Greenberg, Ashton-James, & Ashkanasy, 2007). Finally, onsidering the important influene that role theory has played on betweenperson OCB researh, it ould also be applied to extend our understanding of withinperson flutuations in OCB. At the between-person level, sholars have typially drawn upon role theory to explain why some employees often pereive itizenship as part of their required behavior at work (e.g., Morrison, 1994). Related to our previous disussion about using ESM to study between-person differenes in within-person variability in OCB, role theory is likely to be a useful theoretial perspetive. Speifially, beause role theory suggests that role oneptualizations provide ation templates for employees (Katz & Kahn, 1978), role theory would suggest that an employee is likely to onsistently strive to engage in OCB on a daily basis when she sees OCB as part of her work role (i.e., her daily ation template inludes OCB). Alternatively, when an employee sees OCB as extrarole (i.e., her daily ation template does not inlude OCB), role theory would suggest that daily OCB is likely to flutuate in aordane with the amount of daily resoures available for behaviors that are viewed as not my job (Morrison, 1994, p. 1563). Page 25 of 34

26 Conlusion Although OCB researh has typially foused on identifying anteedents and onsequenes of between-person differenes in levels of OCB, our review douments that researh on within-person flutuations in OCB is beginning to burgeon. Considering the relatively nasent state of the field, we view this domain as one that is ripe for further inquiry. Our review highlights several promising future researh opportunities, and we hope it helps stimulate new and novel researh on the topi. Referenes Alliger, G. M., & Williams, K. J. (1993). Using signal-ontingent experiene sampling methodology to study work in the field: A disussion and illustration examining task pereptions and mood. Personnel Psyhology, 46, Barnes, C. M., Ghumman, S., & Sott, B. A. (2013). Sleep and organizational itizenship behavior: The mediating role of job satisfation. Journal of Oupational Health Psyhology, 18, Barnes, C. M., Hollenbek, J. R., Wagner, D. T., Derue, D. S., Nahrgang, J. D., & Shwind, K. M. (2008). Harmful help: The osts of baking-up behavior in teams. Journal of Applied Psyhology, 93, doi: / Bateman, T. S., & Crant, J. M. (1993). The proative omponent of organizational behavior: A measure and orrelates. Journal of Organizational Behavior, 14, Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tie, D. M. (1998). Ego depletion: Is the ative self a limited resoure? Journal of Personality and Soial Psyhology, 74, Bergeron, D. M. (2007). The potential paradox of organizational itizenship behavior: Good itizens at what ost? Aademy of Management Review, 32, Bergeron, D. M., Shipp, A. J., Rosen, B., & Furst, S. A. (2013). Organizational itizenship behavior and areer outomes: The ost of being a good itizen. Journal of Management, 39, Binnewies, C., Sonnentag, S., & Mojza, E. J. (2009). Daily performane at work: Feeling reovered in the morning as a preditor of day-level job performane. Journal of Organizational Behavior, 30, Page 26 of 34

27 Binnewies, C., Sonnentag, S., & Mojza, E. J. (2010). Reovery during the weekend and flutuations in weekly job performane: A week-level study examining intra-individual relationships. Journal of Oupational and Organizational Psyhology, 83, Bolino, M. C., Klotz, A. C., Turnley, W. H., & Harvey, J. (2013). Exploring the dark side of organizational itizenship behavior. Journal of Organizational Behavior, 34, doi: /job.1847 Bolino, M. C., & Turnley, W. H. (2003). Going the extra mile: Cultivating and managing employee itizenship behavior. The Aademy of Management Exeutive, 17, Cavanaugh, M. A., Boswell, W. R., Roehling, M. V., & Boudreau, J. W. (2000). An empirial examination of self-reported work stress among U.S. managers. Journal of Applied Psyhology, 85, Chang, C., Johnson, R. E., & Yang, L. (2007). Emotional strain and organizational itizenship behaviours: A meta-analysis and review. Work & Stress, 21, Christian, M. S., Eisenkraft, N., & Kapadia, C. (2015). Dynami assoiations between somati omplaints, human energy, and disretionary behaviors: Experienes with pain flutuations at work. Administrative Siene Quarterly, 60, Conway, J. M., Rogelberg, S. G., & Pitts, V. E. (2009). Workplae helping: Interative effets of personality and momentary positive affet. Human Performane, 22, Crant, J. M. (2000). Proative behavior in organizations. Journal of Management, 26, Dalal, R. S. (2005). A meta-analysis of the relationship between organizational itizenship behavior and ounterprodutive behavior. Journal of Applied Psyhology, 90, Dalal, R. S., Lam, H., Weiss, H. M., Welh, E. R., & Hulin, C. L. (2009). A within-person approah to work behavior and performane: Conurrent and lagged itizenshipounterprodutivity assoiations, and dynami relationships with affet and overall job performane. Aademy of Management Journal, 52, Deelstra, J. T., Peeters, M. C. W., Zijlstra, F. R. H., Shaufeli, W. B., Stroebe, W., & Van Doornen, L. P. (2003). Reeiving instrumental support at work: When help is not welome. Journal of Applied Psyhology, 88, doi: / Demerouti, E., Bakker, A. B., Nahreiner, F., & Shaufeli, W. B. (2001). The job demandsresoures model of burnout. Journal of Applied Psyhology, 86, Page 27 of 34

28 Eatough, E. M., Chang, C., Miloslavi, S. A., & Johnson, R. E. (2011). Relationships of role stressors with organizational itizenship behavior: A meta-analysis. Journal of Applied Psyhology, 96, Fay, D., & Sonnentag, S. (2012). Within-person flutuations of proative behavior: How affet and experiened ompetene regulate work behavior. Human Performane, 25, Festinger, L. (1954). A theory of ognitive dissonane. Stanford, CA: Stanford University Press. Fisher, C. D. (2003). Why do lay people believe that satisfation and performane are orrelated? Possible soures of a ommonsense theory. Journal of Organizational Behavior, 24, Fisher, J. D., Nadler, A., & Whither-Alagna, S. (1982). Reipient reations to aid. Psyhologial Bulletin, 91, 27. Fleeson, W. (2001). Toward a struture- and proess-integrated view of personality: Traits as density distributions of states. Journal of Personality and Soial Psyhology, 80, Fleeson, W., & Leiht, C. (2006). On delineating and integrating the study of variability and stability in personality psyhology: Interpersonal trust as illustration. Journal of Researh in Personality, 40, Fox, S., Spetor, P. E., Goh, A., Bruursema, K., & Kessler, S. R. (2012). The deviant itizen: Measuring potential positive relations between ounterprodutive work behaviour and organizational itizenship behaviour. Journal of Oupational and Organizational Psyhology, 85, Fredrikson, B. L. (2001). The role of positive emotions in positive psyhology: The broaden-and-build theory of positive emotions. Amerian Psyhologist, 56, Fritz, C., & Sonnentag, S. (2009). Anteedents of day-level proative behavior: A look at job stressors and positive affet during the workday. Journal of Management, 35, George, J. M. (1991). State or trait: Effets of positive mood on prosoial behaviors at work. Journal of Applied Psyhology, 76, George, J. M., & Brief, A. P. (1992). Feeling good-doing good: A oneptual analysis of the mood at work-organizational spontaneity relationship. Psyhologial Bulletin, 112, Page 28 of 34

29 Glomb, T. M., Bhave, D. P., Miner, A. G., & Wall, M. (2011). Doing good, feeling good: Examining the role of organizational itizenship behaviors in hanging mood. Personnel Psyhology, 64, Goodman, S. A., & Svyantek, D. J. (1999). Person organization fit and ontextual performane: Do shared values matter? Journal of Voational Behavior, 55, Gordon, H. J., Demerouti, E., Bipp, T., & Le Blan, P. M. (2015). The job demands and resoures deision making (JD-R-DM) model. European Journal of Work and Organizational Psyhology, 24, Greenberg, J., Ashton-James, C. E., & Ashkanasy, N. M. (2007). Soial omparison proesses in organizations. Organizational Behavior and Human Deision Proesses, 102, Halbesleben, J. R. B., Neveu, J.-P., Paustian-Underdahl, S. C., & Westman, M. (2014). Getting to the or : Understanding the role of resoures in onservation of resoures theory. Journal of Management, 40, doi: / Halbesleben, J. R. B., & Wheeler, A. R. (2011). I owe you one: Coworker reiproity as a moderator of the day-level exhaustion-performane relationship. Journal of Organizational Behavior, 32, Halbesleben, J. R. B., & Wheeler, A. R. (2015). To invest or not? The role of oworker support and trust in daily reiproal gain spirals of helping behavior. Journal of Management, 41, Higgins, E. T. (1997). Beyond pleasure and pain. Amerian Psyhologist, 52, Hobfoll, S. E. (1989). Conservation of resoures: A new attempt at oneptualizing stress. Amerian Psyhologist, 44, Hulin, C. L., & Judge, T. A. (2003). Job attitudes. In I. B. Weiner (series ed.) & W. C. Borman, D. R. Ilgen, & R. J. Klimoski (vol. eds.), Handbook of psyhology, vol. 12: Industrial and organizational psyhology (pp ). Hoboken, NJ: Wiley. Humphrey, S. E., Morgeson, F. P., & Mannor, M. J. (2009). Developing a theory of the strategi ore of teams: A role omposition model of team performane. Journal of Applied Psyhology, 94, doi: /a Hunter, J. E., & Shmidt, F. L. (2004). Methods of meta-analysis: Correting error and bias in researh findings (2nd ed.). Thousand Oaks, CA: Sage. Page 29 of 34

30 Ilies, R., Sott, B. A., & Judge, T. A. (2006). The interative effets of personal traits and experiened states on intraindividual patterns of itizenship behavior. Aademy of Management Journal, 49, Izard, C. E., & Akerman, B. P. (2000). Motivational, organizational, and regulatory funtions of disrete emotions. In M. Lewis and J. M. Haviland-Jones (Eds.), Handbook of emotion (pp ). New York, NY: Guilford. Johnson, R. E., Lanaj, K., & Barnes, C. M. (2014). The good and bad of being fair: Effets of proedural and interpersonal justie behaviors on regulatory resoures. Journal of Applied Psyhology, 99, Judge, T. A., Simon, L. S., Hurst, C., & Kelley, K. (2014). What I experiened yesterday is who I am today: Relationship of work motivations and behaviors to within-individual variation in the five-fator model of personality. Journal of Applied Psyhology, 99, Katz, D., & Kahn, R. L. (1978). The soial psyhology of organizations. New York, NY: Wiley. Kernis, M. H., Cornell, D. P., Sun, C. R., Berry, A., & Harlow, T. (1993). Theres more to self-esteem than whether it is high or low the importane of stability of self-esteem. Journal of Personality and Soial Psyhology, 65, Klotz, A. C., & Bolino, M. C. (2013). Citizenship and ounterprodutive work behavior: A moral liensing view. Aademy of Management Review, 38, Koopman, J., Lanaj, K., & Sott, B. A. (2016). Integrating the bright and dark sides of OCB: A daily investigation of the benefits and osts of helping others. Aademy of Management Journal, 59, Lanaj, K., Chang, C. H., & Johnson, R. E. (2012). Regulatory fous and work-related outomes: A review and meta-analysis. Psyhologial Bulletin, 138, Larson, R., & Csikszentmihalyi, M. (1983). The experiene sampling method. In H. T. Reis (Ed.), Naturalisti approahes to studying soial interation. San Franiso, CA: Jossey- Bass. Lazarus, R. S., & Folkman, S. F. (1984). Stress, appraisal, and oping. New York, NY: Springer. Lee, K., & Allen, N. J. (2002). Organizational itizenship behavior and workplae deviane: The role of affet and ognitions. Journal of Applied Psyhology, 87, Page 30 of 34

31 LePine, J. A., Erez, A., & Johnson, D. E. (2002). The nature and dimensionality of organizational itizenship behavior: A ritial review and meta-analysis. Journal of Applied Psyhology, 87, Merritt, A. C., Effron, D. A., & Monin, B. (2010). Moral self-liensing: When being good frees us to be bad. Soial and Personality Psyhology Compass, 4, Miner, A. G., & Glomb, T. M. (2010). State mood, task performane, and behavior at work: A within-persons approah. Organizational Behavior and Human Deision Proesses, 112, Miner, A. G., Glomb, T. M., & Hulin, C. (2005). Experiene sampling mood and its orrelates at work. Journal of Oupational and Organizational Psyhology, 78, Mishel, W., & Shoda, Y. (1995). A ognitive-affetive system of personality: Reoneptualizing situations, dispositions, dynamis, and invariane in personality strutures. Psyhologial Review, 102, Morrison, E. W. (1994). Role definitions and organizational itizenship behavior the importane of the employees perspetive. Aademy of Management Journal, 37, Morrison, E. W., & Phelps, C. C. (1999). Taking harge at work: Extra-role efforts to initiate workplae hange. Aademy of Management Journal, 42, Motowidlo, S. J., & van Sotter, J. R. (1994). Evidene that task performane should be distinguished from ontextual performane. Journal of Applied Psyhology, 79, Ohly, S., & Fritz, C. (2010). Work harateristis, hallenge appraisal, reativity, and proative behavior: A multi-level study. Journal of Organizational Behavior, 31, Organ, D. W. (1988). Organizational itizenship behavior: The good soldier syndrome. Lexington, MA: Lexington Books. Organ, D. W., & Ryan, K. (1995). A meta-analyti review of attitudinal and dispositional preditors of organizational itizenship behavior. Personnel Psyhology, 48, Podsakoff, N. P., Whiting, S. W., Podsakoff, P. M., & Blume, B. D. (2009). Individual- and Organizational-level onsequenes of organizational itizenship behaviors: A metaanalysis. Journal of Applied Psyhology, 94, Podsakoff, P. M., MaKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral researh: A ritial review of the literature and reommended remedies. Journal of Applied Psyhology, 88, Page 31 of 34

32 Rodell, J. B., & Judge, T. A. (2009). Can good stressors spark bad behaviors? The mediating role of emotions in links of hallenge and hindrane stressors with itizenship and ounterprodutive behaviors. Journal of Applied Psyhology, 94, Rubin, R. S., Dierdorff, E. C., & Bahrah, D. G. (2013). Boundaries of itizenship behavior: Curvilinearity and ontext in the itizenship and task performane relationship. Personnel Psyhology, 66, doi: /peps Salovey, P., Mayer, J. D., Goldman, S. L., Turvey, C., & Palfai, T. P. (1995). Emotional attention, larity, and repair: Exploring emotional intelligene using the Trait Meta-Mood Sale. In J. W. Pennebaker (Ed.), Emotion, dislosure, and health (pp ). Washington, DC: Amerian Psyhologial Assoiation. Shreurs, B. H. J., Van Emmerik, I. J., Gunter, H., & Germeys, F. (2012). A weekly diary study on the buffering role of soial support in the relationship between job inseurity and employee performane. Human Resoure Management, 51, Sott, B. A., Barnes, C. M., & Wagner, D. T. (2012). Chameleoni or onsistent: A multilevel investigation of emotional labor variability. Aademy of Management Journal, 55, Shokley, K. M., Ispas, D., Rossi, M. E., & Levine, E. L. (2012). A meta-analyti investigation of the relationship between state affet, disrete emotions, and job performane. Human Performane, 25, Sonnentag, S. (2003). Reovery, work engagement, and proative behavior: A new look at the interfae between nonwork and work. Journal of Applied Psyhology, 3, Sonnentag, S., & Starzyk, A. (2015). Pereived prosoial impat, pereived situational onstraints, and proative work behavior: Looking at two distint affetive pathways. Journal of Organizational Behavior, 36, Spetor, P. E., & Fox, S. (2002). An emotion-entered model of voluntary work behavior: Some parallels between ounterprodutive work behavior and organizational itizenship behavior. Human Resoure Management Review, 12, Spene, J. R., Brown, D. J., Keeping, L. M., & Lian, H. (2014). Helpful today, but not tomorrow? Feeling grateful as a preditor of daily organizational itizenship behaviors. Personnel Psyhology, 67, Spene, J. R., Ferris, D. L., Brown, D. J., & Heller, D. (2011). Understanding daily itizenship behaviors: A soial omparison perspetive. Journal of Organizational Behavior, 32, Page 32 of 34

33 Smith, C. A., Organ, D. W., & Near, J. P. (1983). Organizational itizenship behavior: Its nature and anteedents. Journal of Applied Psyhology, 68, Trougakos, J. P., Beal, D. J., Cheng, B. H., Hideg, I., & Zweig, D. (2015). Too drained to help: A resoure depletion perspetive on daily interpersonal itizenship behaviors. Journal of Applied Psyhology, 100, Van Dyne, L., & Ellis, J. B. (2004). Job reep: A reatane theory perspetive on ob as overfulfillment of obligations. In J. C. Shapiro, L. M. Shore, S. Taylor, & L. E. Tetrik (Eds.), The employment relationship: Examining psyhologial and ontextual perspetives (pp ). Oxford, England: Oxford University Press. Van Dyne, L., & LePine, J. A. (1998). Helping and voie extra-role behaviors: Evidene of onstrut and preditive validity. Aademy of Management Journal, 37, Vidyarthi, P. R., Liden, R. C., Anand, S., Erdogan, B., & Ghosh, S. (2010). Where do i stand? Examining the effets of leader-member exhange soial omparison on employee work behaviors. Journal of Applied Psyhology, 95, Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affet: The PANAS sales. Journal of Personality and Soial Psyhology, 54, Weiss, H. M., & Cropanzano, R. (1996). Affetive events theory: A theoretial disussion of the struture, auses, and onsequenes of affetive experienes at work. Researh in Organizational Behavior, 18, Williams, L. J., & Anderson, S. E. (1991). Job satisfation and organizational ommitment as preditors of organizational itizenship and in-role behaviors. Journal of Management, 17, Wood, J. V. (1996). What is soial omparison and how should we study it? Personality and Soial Psyhology Bulletin, 22, Wu, C., & Parker, S. K. (2012). The role of attahment styles in shaping proative behaviour: An intra-individual analysis. Journal of Oupational and Organizational Psyhology, 85, Xanthopoulou, D., Heuven, E., Demerouti, E., Baker, A. B., & Shaufeli, W. (2008). Working in the sky: A diary study on work engagement among flight attendants. Journal of Oupational Health Psyhology, 13, Page 33 of 34

34 Brent A. Sott Broad College of Business, Department of Management, Mihigan State University Fadel K. Matta University of Georgia Joel Koopman Department of Management, University of Cininnati Page 34 of 34