Examining correlates of organizational citizenship behavior and counterproductive work behavior in a collectivist culture: the case of Arab teachers in Israel

Purpose – This study aims to examine the relationships between personal (emotional intelligence, Dark Triad (DT), core self-evaluation and burnout) and situational variables (organizational justice) and organizational citizenship behavior (OCB) (supervisor report) and counterproductive work behavior (CWB) (self-report). Design/methodology/approach – In total, 680 questionnaires were distributed to teachers in 20 Arab elementary schools in Northern Israel. Usable questionnaires were returned by 509 teachers (75%). The questionnaires covered emotional intelligence, DT, core self-evaluation, organizational justice, burnout, CWB and demographic characteristics. Their principals filled out questionnaires on the teachers’ in-role performance and OCB. Findings – Results showed that CWB was mostly related to higher levels of psychopathy, lower levels of emotional intelligence (ability to use emotions) and higher levels of burnout (emotional exhaustion). OCB was related to higher levels of procedural justice, lower levels of burnout and higher levels of emotional intelligence. Practical implications – Organizations should consider ways to reduce burnout, which may reduce CWB and increase perceptions of justice, thereby promoting OCB. Originality/value – Two novel aspects are noteworthy. First, this study simultaneously examines both CWB and OCB to clarify the similarities and differences between them. Second, few studies have examined the correlates of CWB and OCB in Arab culture.


Introduction
Two employee workplace behaviors have attracted much attention in recent years, namely, organizational citizenship behavior (OCB) and counterproductive work behavior (CWB).
Both are spontaneous and voluntary and have significant effects on organizational success and failure (Berry, Ones, & Sackett, 2007;Organ, Podsakoff, & MacKenzie, 2006). OCB refers to organizationally valuable behaviors that cannot be imposed by formal role responsibilities or evoked by a contractual assurance of reimbursement; it includes both impersonal OCB (directed toward the organization in general) and altruistic OCB (helping a specific person within the organization; Organ (1988)). CWB refers to deliberate actions that damage the organization or its members (O'Boyle, Humphrey, Pollack, Hawver, & Story, 2011). It comprises negative behaviors that can be directed toward organizations (CWB-O) or individuals (CWB-P). CWB is considered one of the most damaging behaviors within organizations (Cohen, 2016).
The OCB and CWB variables have been highly researched in their relevant fields (Miles, Borman, Spector, & Fox, 2002). Thus, there is a significant amount of research on their correlates; however, two aspects have not received enough attention in the literature, namely, first, a simultaneous examination of both CWB and OCB to clarify their similarities and differences, which is necessary, as recent literature has contended the two are distinct constructs, rather than a single continuum (Dalal & Carpenter, 2018;Sackett, Berry, Wiemann, & Laczo, 2006). Second, few studies have examined the correlates of CWB and OCB within a non-Western culture. Arab culture is regarded as traditional and collectivist compared to modern and individualistic Western cultures (Cohen, 1999).
This study aims to contribute to the research trends suggested above by examining the relationships between CWB and OCB and personal and situational factors among Arab-Israeli teachers in Israel. It is noteworthy that few studies have examined these relationships among teachers or other helping professionals in non-Western societies. This research thus, offers several important and novel contributions. First, the study proposed and tested a comprehensive model explicating the relationships between personal and situational variables, and OCB and CWB. Second, this research examined determinants of CWB and OCB in a traditional Arab setting and culture. Most literature investigating these behaviors has concentrated on Western cultures; few studies have examined this issue in non-Western cultures such as China (Liu & Cohen, 2018), Philippines (Robertson, Datu, Brawley, Pury, & Mateo, 2016) and Turkey (Jonason, Okan, & Özsoy, 2019). Examining these relationships within a traditional collectivist culture is important to evaluate whether explanations for CWB and OCB are similar to those in more individualistic Western cultures (Minkov, 2011). Third, unlike most previous studies, this study examined both CWB and OCB simultaneously, as dependent variables, which is critical, as scholars currently posit that OCB and CWB do not represent opposite behaviors but independent constructs (Dalal & Carpenter, 2018;Sackett et al., 2006). Therefore, exploring the different mechanisms affecting each behavior is OMJ important. Finally, this study examined data from different types of reports (principals' reports for OCB and teachers' self-reports for CWB), which increases the validity of the findings.

Study setting
The hypotheses presented here are specifically based on Arab teachers employed in Arab elementary schools, operating in Arab-populated cities or villages in Israel. Arabs represent about one-sixth of Israel's population, are a permanent and non-assimilating minority, are clearly distinguished from Jews in places of residence, culture and language and adhere to their own traditions (Cohen & Abedallah, 2015;Cohen & Kirchmeyer, 1995). Israeli Arabs have a traditional collectivist culture. The collectivist orientation is expressed in ideals such as solidarity, cooperation, commitment, mutual trust, support and a sense of belonging that are believed to be present in Arab nuclear and extended families and in the community (Pines & Zaidman, 2003;Ronen & Shenkar, 1985).
Israel's 7 million citizens include 1.5 million students in the educational system, of whom 25% are Arabs. Arabic schools are segregated in Israel, and offer a curriculum that emphasizes Arabic history, religion and culture. The Arabic education system consists of four tiers ranging from kindergarten to grade 12: (1) Pre-elementary education (kindergarten).
Compulsory education ranges from kindergarten to 10th grade, although most students complete the 12th grade. In 2018-2019, female schoolteachers comprised most (73.9%) of the Arab sector's educators (Israel Central Bureau of Statistics, 2019). As for Arab school teachers' educational level, 3.3% have non-academic degrees (7.2% in the Jewish sector); 66.6% have a bachelor's degree (56.1% in the Jewish sector); and 31.6% have a master's degree or higher (36 per cent in the Jewish sector; Israel Central Bureau of Statistics 2019).

Conceptual framework and research hypotheses
The research model is presented in Figure 1. In the following sections, the theory and hypotheses regarding the relationships in the model are presented.
Emotional intelligence, in-role performance and organizational citizenship behavior Emotional intelligence has had a significant impact on managerial practice (Ashkanasy & Daus, 2002), mainly because it has the potential to contribute to more positive attitudes, behaviors and outcomes among organizational members (Carmeli, 2003). This study adopted Salovey and Mayer's (1990, p. 189) definition of emotional intelligence, "the ability to monitor one's own and others' feelings and emotions, to discriminate among them and to use this information to guide one's thinking and actions." They identified three dimensions of emotional intelligence, which were used in this study: (1) The ability to appraise and express emotions in oneself and others.
(2) The ability to regulate emotions in oneself and others.
Emotional intelligence is related to in-role performance because organizational settings require interpersonal interaction. Most interpersonal interactions are related to the Work behavior in a collectivist culture performance of job duties. To facilitate effective interactions, which are essential to job performance, particularly among teachers, it is essential that individuals are able to understand and manage their own emotions and those of others (Carmeli & Josman, 2006;Wong & Law, 2002). Two meta-analyzes (Joseph & Newman, 2010;O'Boyle, Forsyth, & O'Boyle, 2011) supported the relationship between emotional intelligence and job performance.
Regarding OCB, individuals with high emotional intelligence may be more inclined to identify and respond appropriately to the emotions of coworkers, customers and superiors (Day and Carroll, 2004). Because they are emotionally perceptive, they are likely to handle emotionally laden situations in ways that exceed their job description or organizational rules. As they are alert to others' feelings, they may help manage interactions within their workgroups. Emotionally intelligent employees are likely to be empathetic toward the organization, enabling them to adopt the organization's perspective and behave in a manner that will benefit the organization (Cohen & Abedallah, 2015). Carmeli (2003) and Carmeli and Josman (2006) argued that emotional intelligence may enhance altruistic behavior, as emotional intelligence enables employees to shift easily from negative to positive moods, and employees with positive emotions are more likely to engage in helpful behaviors. Additionally, these authors argued that involvement in altruistic behavior is rewarding for emotionally intelligent employees because it maintains their positive state of mind. The findings presented by Carmeli (2003), Carmeli and Josman (2006) and by Hemmati, Rezapur, Hashemi, and Mohammadi (2013), support these contentions. Thus, the first hypothesis (H1) of this study states: H1. Emotional intelligence will be positively related to in-role performance and OCB.

OMJ
Emotional intelligence and counterproductive work behavior Emotional intelligence has an important role in preventing negative behaviors (Martin, Knopoff, & Beckham, 1998). Employees with lower levels of emotional intelligence are incapable of withholding negative emotions that result from negative experiences in the workplace, and may behave aggressively and inappropriately in a manner that damages the organization and/or its employees (Jung & Yoon, 2012;Quebbeman & Rozell, 2002). Contrastingly, employees with higher levels of emotional intelligence might respond to negative work experiences more effectively, demonstrating an ability to effectively control and manage their emotions (Greenidge, Devonish, & Alleyne, 2014).
To summarize, employees who cannot control their emotions may fail in their social interactions and experience more negative emotions that will lead to more CWB (Jung & Yoon, 2012). A negative relationship exists between emotional intelligence and CWB (Deshpande, Joseph, & Shu, 2005;Trinidad & Johnson, 2002). Thus, the second hypothesis (H2) of this study states: H2. Emotional intelligence will be negatively related to CWB.

Dark triad, in-role performance and organizational citizenship behavior
The DT consists of three anti-social subclinical personality traits, namely, Machs (manipulation of others), narcissism (feelings of grandiosity, entitlement, dominance and self-superiority) and psychopathy (high impulsivity and thrill seeking, low empathy and anxiety). The three personality traits are overlapping but distinct constructs (Stead, Fekken, Kay, & McDermott, 2012;Lyons, 2019). Paulhus and Williams (2002) Stead et al. (2012) argued that these traits share a socially malevolent character with self-promoting behavioral tendencies, emotional coldness, duplicity and aggressiveness. Individuals who possess this personality style tend to carry out a disproportionately large amount of anti-social behavior on others. Therefore, identifying the correct way to conceptualize these traits, and to classify them within the broader context of psychopathology, will ultimately lead to diagnosing and treating individuals with the DT personality style (Stead et al., 2012).
Because OCB is a social and discretionary behavior that prioritizes others and the group over the self, it is probable that individuals with high DT traits will be less likely to engage in OCB. Furthermore, people high in Machs focus on manipulation and prioritize themselves at the expense of others, so it is less likely that Machs would engage in OCB. Those high in narcissism (Narcissists) have a strong sense of entitlement and their high self-evaluation may preclude them from engaging in discretionary helping behaviors such as OCB. Psychopaths' lack of concern for others also makes it less likely for them to engage in discretionary helping behaviors that promote the welfare of the organization and their coworkers (Webster & Smith, 2019).
According to Bourdage, Lee, Lee, and Shin (2012), while the relationship between psychopathy and OCB is expected to be negative, the relationship between the other two dark personality traits and OCB may not be as straightforward. However, certain factors such as impression management, long-term orientation and the use of soft tactics may prompt Machs and Narcissists to display some OCBs in certain situations (Bourdage et al., 2012;Lyons, 2019). Bourdage et al. (2012) found that subclinical psychopathy has a negative effect on OCB and in-role performance, and that Machs and subclinical narcissism were nonsignificant predictors of OCBs. Thus, the third hypothesis (H3) of this study states: Work behavior in a collectivist culture H3: Psychopathy will be negatively related to in-role performance and OCB. No relationship will be found between Machs and Narcissism with in-role performance and OCB.

Dark triad and counterproductive work behavior
There is a simple justification for expecting that the DT will be related to CWB: Deviant workplace behaviors may be best predicted by deviant personality traits (Cohen, 2016). Specifically, psychopaths believe they are above the social, moral, ethical and legal principles that govern society and rarely experience shame, guilt, remorse or regret. Thus, such beliefs may easily lead them to perform CWBs. Furthermore, narcissism is expected to be positively related to CWBs for at least two reasons, namely, first, because narcissists view themselves as highly important, they are willing to violate rules for personal gain. Second, narcissism overlaps, conceptually and empirically, with impulsiveness (Cohen, 2018); as CWBs are often performed impulsively, narcissism can be expected to have a positive relationship with CWB. Moreover, individuals rating high in Machiavellianism (high Machs) are likely to engage in highly manipulative CWBs when facing impediments to achieving their goals, are remorseless, willingly engage in hostile and unethical behaviors (Liu & Cohen, 2018;Lyons, 2019), and are prone to making unethical decisions and often assume that others would make the same choices. They are more likely to lie to, steal from, cheat and mislead others (Cohen, 2016(Cohen, , 2018. Thus, the fourth hypothesis (H4) of this study states: H4. The DT will be positively related to CWBs.
Core self-evaluation, in-role performance and organizational citizenship behavior Core self-evaluation refers to fundamental and broad evaluations of self-regulatory capacities (Johnson, Rosen, & Levy, 2008). Specifically, it refers to individuals' fundamental beliefs about themselves and their self-worth (Judge, Locke, & Durham, 1997). Judge, Erez, Bono, and Thoresen (2003) contended that core self-evaluation involves evaluations of selfworth, control over one's environment, capability and competence to be successful and emotional adjustment. Self-esteem is the most representative of its sub-dimensions; it is a core concept that refers to an individual's ability to comprehensively summarize his or her self-worth. Workers with high self-esteem are more willing to accept challenging work and maintain a positive and optimistic outlook regardless of their work outcomes (Judge, Erez, & Bono, 1998).
A positive core self-concept affects an individual's general level of initiative and beliefs concerning hihe/sher general level of competence. As OCB requires individuals to initiate social interactions and to be confident about their interpersonal capabilities, those who have high core self-evaluations are prone to engage in OCB that requires initiative (Bowling et al., 2012;Ferris, Rosen, Johnson, Brown, Risavy, & Heller, 2011;Somech & Drach-Zahavy, 2000). Thus, the fifth hypothesis (H5) of this study states: H5. High core self-evaluation will be related to high in-role performance and OCB.
Core self-evaluation and counterproductive work behavior A negative relation between self-esteem and CWB is proposed by consistency theory (Whelpley & McDaniel, 2016). Within this context, individuals are motivated to believe and OMJ act in ways that run parallel with their self-views. Consistency theory predicts lower levels of CWB among individuals with high self-esteem for two reasons, namely, first, individuals with high self-esteem are motivated to perform their jobs well, and CWB is considered in job performance ratings; thus, people attempting to maximize their performance are less likely to engage in CWB. Second, as these individuals view themselves positively, they tend not to engage in CWB to avoid making negative judgments about themselves. Accordingly, consistency theory predicts that individuals with high self-esteem will engage in fewer CWBs, and individuals with lower self-esteem will engage in more CWBs (Whelpley & McDaniel, 2016). Thus, the sixth hypothesis (H6) of this study states: H6. High core self-evaluation will be related to low CWB.
Organizational justice, in-role performance and organizational citizenship behavior Two sources of organizational justice are cited frequently in the literature (Cohen, 2016;Rosen, Chang, Johnson, & Levy, 2009). The first, distributive justice, addresses the fairness of outcomes, including outcomes such as office assignments, promotions, job titles and the like (Karriker & Williams, 2009). Drawing from equity theory, an individual who experiences inequity or injustice will attempt to restore balance using a number of mechanisms, including reduction of in-role performance and OCB (Ang, Van Dyne, & Begley, 2003;Devonish & Greenidge, 2010;Greenberg, 1990).
The second, procedural justice, has two dimensions. The first dimension is formal procedures and refers to the perceived fairness of the procedures associated with outcome distributions. The second dimension, interactional justice, refers to the perceived fairness of interpersonal treatment from those administering the procedures used to arrive at certain outcomes (Devonish & Greenidge, 2010). When organizational decision-making is consistent and meets the bias suppression rule (uniform treatment of all), employees assess procedural justice positively (Ang et al., 2003;Colquitt, 2001;Greenberg, 1990), leading to higher levels of performance. Perceptions of procedural justice affect an employee's general perception of whether an organization values him or her. This perception may influence the employee's OCB by prompting him/her to define hihe/sher relationship with the organization as one of social exchange (Moorman, Blakely, & Niehoff, 1998). Furthermore, the quality of the managers' personal treatment of the employee may influence the latter's personal gestures (Cohen-Charash & Spector, 2001). Thus, the seventh hypothesis (H7) of this study states: H7. Organizational justice is positively related to in-role performance and OCB.
Organizational justice and counterproductive work behavior Cohen-Charash and Spector (Cohen & Diamant, 2017) contended that CWBs could be seen as reactions to perceived injustice and are primarily related to an employee changing hihe/sher input to restore equity. When employees perceive distributive injustice, they may damage the organization to make the outcome/input ratio less negative from their perspective. Furthermore, perceived injustice leads to negative perceptions of the organization, and, hence, to CWBs (Hershcovis et al., 2007;Liu & Berry, 2013). Crawshaw, Cropanzano, Bell, and Nadisic (2013) contended that individuals who act assertively to restore fairness frequently engage in retributive behavior: mistreatment and perceived injustice lead to moral outrage and the desire to punish the perpetrators. Thus, employees may respond to unfairness by engaging in CWBs, such as organizational deviance, sabotage and aggression. Martinson, Anderson, Crain, and De Vries (2006) contended that perceptions of injustice Work behavior in a collectivist culture may threaten an individual's identification with a group, leading to compensatory behaviors such as CWBs. Thus, the eighth hypothesis (H8) of this study states: H8. Organizational justice is negatively related to CWBs.
Burnout, in-role performance and organizational citizenship behavior Burnout is a state of mental and emotional exhaustion that is characterized by feelings of being drained by others. It is defined by three dimensions, namely, emotional exhaustion, depersonalization and reduced personal accomplishment (Emmerik, Jawahar, & Stone, 2005;Sesen, Çetin, & Basim, 2011). The relationship between burnout and job performance may best be explained by the conservation of resources (COR) model of stress (Hobfoll & Freedy, 1993).
The COR model suggests that burnout occurs in response to the loss or perceived loss of resources; employees then take actions to protect themselves, such as putting less effort into their work, resulting in decreased job performance (Halbesleben & Bowler, 2005) and reduced OCB. Taris (2006), also following the COR model, argues that high levels of burnout signify that workers possess insufficient resources to effectively deal with the demands of their jobs. Burnout is often accompanied by feelings of depression and loss of self-esteem, physical components such as hypertension and behavioral aspects including alcoholism and drug use (D'Amato & Zijlstra, 2008). All these factors have a detrimental effect on work outcomes. Studies have shown a negative relationship between burnout and in-role performance (Cropanzano, Rupp, & Byrne, 2003;Halbesleben & Bowler, 2005;Parker & Kulik, 1995).
To the extent that burnout resulting from chronic long-term stress can negatively affect variables associated with intrinsic motivation (e.g. engagement with work and sense of achievement; Maslach (1986)), it is appropriate to assume that feelings of exhaustion, depersonalization (i.e. a callous and detached response to job duties) and a diminished sense of professional achievement exert a negative effect on employees' willingness to put extra effort into their work (Cropanzano et al., 2003;Emmerik et al., 2005). Less OCB from emotionally exhausted employees is, therefore, to be expected. Thus, the ninth hypothesis (H9) of this study states: H9. Burnout will be negatively related to in-role performance and OCB.

Burnout and counterproductive work behavior
The relationship between burnout and CWB can also be explained by the COR model of stress (Hobfoll & Freedy, 1993). Here, burnout also happens in response to the loss or threatened loss of resources, so employees tend to take actions to protect them (Halbesleben & Bowler, 2005). Following the COR model, high levels of burnout signify that workers have deficient resources to deal with the demands of their jobs. The result is feelings of depression and loss of self-esteem, and sometimes hypertension (Taris, 2006;D'Amato & Zijlstra, 2008). All these factors have a damaging effect on work outcomes and might lead to OCB. A growing body of studies suggests that burnout can lead to CWBs (Banks, Whelpley, Oh, & Shin, 2012;Fox et al., 2001;Hershcovis et al., 2007;Salami, 2010). The feeling of exhaustion leads to employees blaming the organization for their exhaustion because of value incongruence. This may result in negative attitudes toward the organization and involvement in CWBs (Ansari et al., 2013). Thus, the 10th hypothesis (H10) of this study states: H10. Burnout is positively related to CWBs.

Subjects and procedures
This study examined Arab teachers working in elementary schools in northern Israel, where 44% of the Arab population of Israel reside in Arab communities. The teachers completed questionnaires on emotional intelligence, DT, core self-evaluation, organizational justice, burnout, CWB and demographic characteristics. Their principals completed questionnaires addressing the teachers' in-role performances and OCBs, in most cases one or two months following the collection of questionnaires from teachers. In total, 680 questionnaires were distributed to teachers across 20 Arab elementary schools.
Formal permission to proceed with the survey was obtained from the Israeli Ministry of Education. In addition, informed consent was obtained from the respondents. Usable questionnaires were returned by 509 teachers (75%). The teachers indicated their national identity numbers on the questionnaires to allow us to match their responses with the principals' evaluations. The questionnaires were translated from English to Arabic, using the common process of translation and back-translation by speakers of Arabic and English, and were administered on-site and took about 20 min to complete. No compensation was provided.
Women comprised 77.8% of the participating teachers, and the average age of participants was 40.27. The average tenures in the school and in the profession were 11.5 years and 16.6 years, respectively. Married participant's comprising 91.7%, 52.3% possessed a bachelor's degree and 14.9% had a master's degree. Table 1 presents the basic information regarding the study scales.

Scales
All items were measured on a scale from one to seven. In addition, three demographic variables were used as control variables, namely, age (ratio), gender (1 = male; 2 = female) and education (ordinal). Table 2 presents the descriptive statistics of the research variables and the intercorrelations between them. The results revealed acceptable reliability of the variables. Because of the very high correlation between distributive justice and procedural justice (0.075; p < 0.001) it was decided to combine the formal procedure scale and the interactional justice scale into one scale presenting procedural justice. As for the remaining correlations, none of them (except the relationship between the two dimensions of OCB) exceeded 0.60, thus reducing the possibility of multi-collinearity.

Results
Additionally, the study performed several confirmatory factor analyzes of the research variables (Table 3). First, the paper compared the fit of a three-factor model for emotional intelligence to the fit of an alternate one-factor model. The results revealed superior fit indices for the three-factor model than those found for the one-factor model. Second, the fit of a three-factor model for DT was compared to the fit of an alternate one-factor model. Again, the results for the three-factor model revealed a superior fit. Similar results were discovered for the superiority of a two-factor model for organizational justice, a three-factor model for burnout, a two-factor model for CWB and a three-factor model for OCB (Table 3). These results are consistent with the absence of common method variance.
Emotional intelligence, in-role performance organizational citizenship behavior and counterproductive work behavior Hierarchical linear modeling (HLM) analyzes were performed for each of the dependent variables (three OCB and in-role variables and two CWB variables). OMJ results of the HLM analysis. H1 was meagerly and partially supported: only one dimension of emotional intelligence, the ability to express emotions (Line 4), was related to individual OCB (positively). H2 was partially supported: the ability to use emotions (Line 6) was negatively related to CWB.
Dark triad, in-role performance, organizational citizenship behavior and counterproductive work behavior H3 was not supported: psychopathy (Line 9) was not related to in-role performance and OCB and Machs (Line 7) alone was negatively related to in-role performance. H4 was partially supported: psychopathy (Line 9) was significantly related to CWB (positively).
Core self-evaluation, in-role performance, organizational citizenship behavior and counterproductive work behavior H5 was not supported: no relationship was found between core self-evaluation (Line 10), inrole performance and OCB. H6 was partially supported: core self-evaluation (Line 10) was negatively related to CWB (Table 4).
Organizational justice, in-role performance, organizational citizenship behavior and counterproductive work behavior H7 was partially supported: procedural justice (Line 15) was positively related to individual and organizational OCB. H8 was not supported: distributive justice (Line 14) was positively related to individual CWB, contrary to the hypothesis.
Work behavior in a collectivist culture of burnout (Lines 11-13) were significantly related positively to the two dimensions of CWB (except personal accomplishment for individual CWB). Finally, the demographic control variables showed some noteworthy relationships. First, gender had a strong relationship with CWB, revealing that men perform CWB more than women. Interestingly, younger teachers were discovered to perform more individual CWB, while older teachers perform more individual OCB. Those with less education were found to perform more organizational CWB, while those with higher education performed more individual OCB.

Discussion
This study examined the relationship between personal and situational variables with two important work outcomes, namely, CWB and OCB. The relationships were examined in an Arab collectivist culture. One expectation of this study was that DT would have a weaker effect on the two work outcomes because of the social control mechanisms inherent in a collectivist culture. The findings provide reasonable support for this expectation. Psychopathy was positively related to the two dimensions of CWB; however, no relationship was discovered between Machs or narcissism and CWB. Additionally, although there was a negative relationship between Machs and in-role performance, no other relationships were found between the DT, OCB and in-role performance. Grijalva andNewman (2015), following O'Boyle, Forsyth, Banks, andMcDaniel (2012), explain these weak relationships by contending that cultures with high in-group collectivism (e.g. the Arab population examined here) suppress narcissistic expressions such as CWBs collectivist cultural cues may suggest to narcissists that there will be harsher sanctions for individuals who violate group norms and harm a group or organization, motivating them to refrain from performing CWBs. However, two issues remain unanswered by the above analysis, which future research should address. First, why was the above explanation not relevant for psychopaths? Is it because this trait affects CWB above and beyond the effects of culture? Second, there was a lack of any measurable relationship between the DT and OCB; this finding supports the contention that CWB and OCB are different, but not opposite constructs, considering the positive relationship between psychopathy and CWB. Results showed a moderate to weak effect of emotional intelligence and core selfevaluation on CWB, OCB and in-role performance. Regarding emotional intelligence, only one dimension (ability to use emotions) was negatively related to the two dimensions of CWB. This finding provides some support for the study by Winkel, Wyland, Shaffer, and Clason (2011), who argued that emotionally intelligent individuals can "read" social situations in the workplace and identify opportunities to assist others. Here, emotional intelligence helped individuals identify situations in which they should refrain from behaviors that could harm their organization or coworkers. This finding highlights emotional intelligence as a concept that warrants further examination in its relationship to Notes: *p # 0.05; **p # 0.01; ***p # 0.001. Gender: 1 = male; 2 = female Work behavior in a collectivist culture work performance outcomes. Moreover, only one significant relationship was found between core self-evaluation, OCB and CWB. The expected negative relationship between core self-evaluation and organizational CWB showed that core self-evaluation may have a preventive effect against performing CWB. Future research should examine core selfevaluation and its relationship to OCB and CWB. One of the most important results of this study is the relationship of burnout, OCB, inrole performance and CWB. Two dimensions of burnout were positively related to the two dimensions of CWB. One dimension of burnout (emotional exhaustion) was negatively related to individual OCB and in-role performance, as expected. The strong relationship between burnout, individual OCB and in-role performance confirms the findings of Liu, Zhou, and Che (2019), and provides strong support for the COR model of stress (Hobfoll & Freedy, 1993). According to this model, burnout occurs in response to either the loss or perceived loss of resources, and one way by which employees may seek to protect resources is to put less effort into their work, resulting in lower job performance (Halbesleben & Bowler, 2005) and OCB.
The positive relationship between burnout and CWB supports the significant positive relationship between the two reported by Shkoler and Tziner (2017), who showed that employees who feel an emotional imbalance may try to create a sense of balance and reduce negative feelings by exercising coping strategies, including CWB. Koon and Pun (2018) found a positive relationship between emotional exhaustion and workplace incivility; excessive job demands give rise to emotional exhaustion, which leads to workplace incivility. This implies that organizations should take preventative steps to alleviate employees' emotional exhaustion by reducing the pressure of job demands wherever possible.
The positive relationship between procedural justice and the two dimensions of OCB support the meta-analysis of Cohen-Charash and Spector (2001), who explained that, as employees perceive procedural justice to be a part of organizational conduct and, as they want to keep their work organization just, they will contribute to the betterment of their fellow employees by contributing more than their role demands. The positive relationship between distributive justice and individual CWB was not expected and can be related to the perceptions of Arab teachers as deprived minorities (Cohen, 1999). This assumption is strengthened by the finding of no relationship between distributive justice and CWB in a sample of Jewish teachers in Israel (Cohen & Diamant, 2017).
The findings of this study showed significant relationships between the demographic variables and CWB, for example, strong negative relationships between gender and the two dimensions of CWB, strong negative relationships between age and individual CWB and negative relationships between education and organizational CWB. These relatively strong relationships challenge the very weak relationship between demographic variables and CWB found in the meta-analysis of Berry et al. (2007). In addition, this study found significant relationships between the demographic variables and OCB. Age and education were positively related to individual OCB. Stronger relationships between demographic variables and OCB were found by Abd El Majid and Cohen (2015), also among Arab teachers. This finding should also be compared to the very weak relationship between age and education and OCB found in the meta-analysis of Carpenter, Berry, and Houston (2014). These findings underscore the possibility that demographic variables have a stronger effect on CWB and OCB in collectivist cultures than in individualistic ones. This possibility should be tested in future studies.

Practical implications
The current study has several practical implications. First, results suggest that organizations should pay more attention to the role of burnout in relation to OCB and CWB. For example, the organizational climate of respect is inversely related to burnout; developing a positive organizational climate can reduce burnout and increase OCB (Osatuke, Moore, Ward, Dyrenforth, & Belton, 2009). Managers should be aware that some of these strategies may increase employees' workload, which, in turn, increase the likelihood of burnout, thereby increasing CWB, and reducing OCB and possibly, job performance. The above pattern was mentioned by Koon and Pun (2018), who found a positive relationship between emotional exhaustion and workplace incivility. Koon and Pun (2018) explained that excessive job demands give rise to emotional exhaustion, which, instigates workplace incivility. This implies that organizations should take preventative steps in alleviating employees' emotional exhaustion by reducing the pressure of job demands (Koon and Pun, 2018). Organizations that want to encourage more OCB and less CWB should not overload employees or deplete their resources (Cohen & Abedallah, 2015;Shkoler & Tziner, 2017).
Second, the positive relationship between procedural justice and the two dimensions of OCB support the findings of Cohen-Charash and Spector (2001), who, based on their metaanalysis, explained that, as employees perceive procedural justice to be a part of organizational conduct, and, as they desire a just work organization, they will be willing to contribute to the betterment of their organization and fellow employees by spontaneous and informal activities. This implies that organizations and managers should be aware of and sensitive to the perceptions of their employees of the fairness of organizational procedures. Organizations and managers should better communicate the procedures in the organization and the effort made by the organization to make just procedures, which requires organizational transparency (Cohen-Charash & Spector, 2001).
Third, the weak effect of DT on OCB and CWB suggests that the ethnic composition of employees should be considered by organizations and managers in managing dark personalities. Based on these findings and those of Liu and Cohen (2018) in a Chinese context, it can be concluded that collectivist cultures that promote the advancement of the group rather than the individual may provide stronger environmental constraints against DT than more individualistic cultures. Behaviors that constitute DT traits might not be socially desirable in non-Western societies because in collectivist cultures individuals have a strong sense of duty to family and others that form their societal in-group (Robertson et al., 2016). This finding should be considered when managing dark personality traits in the workplace.

Limitations
Two limitations are noteworthy. First, because one of the independent variables was collected from only one source, the likelihood of common method errors exists. Second, this study examined one occupation in one culture, so caution is advised before generalizing the findings. However, despite these limitations, the findings have important implications that can inform both future research and the practice of organizational management. Chief among these is the conclusion that organizations that depend upon employees' extra-role contributions to function effectively must protect them from resource depletion and burnout.

Declaration of interest statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Work behavior in a collectivist culture