Introduction

Ostracism—being ignored and excluded by others—is a universal phenomenon in the workplace (Ferris et al., 2008; Williams, 2009). A survey by Fox and Stallworth (2005) revealed that 66% of employees are given silent treatment or ignored over a 5-year period. Such ostracism is a painful and threatening experience, and those who are ostracized experience dysfunctional consequences at work (for details, see the meta-analyses of Howard et al. [2020] and Li et al. [2021]) and at home (Liu et al., 2013; Zhang et al., 2023). Because a better understanding of the predictors of ostracism can minimize exposure to workplace ostracism (Wu et al., 2021), it is essential to investigate why and how ostracism occurs (Howard et al., 2020; Rudert et al., 2020). Studies have shown that individual differences (e.g., the need to belong), deviant behaviors (e.g., incivility, unethical behavior), job characteristics (e.g., goal interdependence), and environmental factors (e.g., leadership, job climate) can influence ostracism in the workplace (see, e.g., Howard et al., 2020; Kanwal et al., 2019; Kwan et al., 2022; Wu et al., 2015).

Despite these fruitful findings, the target personalities that predict the occurrence of ostracism remain poorly understood. Two studies explored the effect of the Big Five on workplace ostracism. Wu et al. (2011) discovered that extraversion, agreeableness, and emotional stability were negatively associated with workplace ostracism, while conscientiousness and openness to experience were unrelated to workplace ostracism. Rudert et al. (2020) found that individuals low in agreeableness or conscientiousness were more likely to be ostracized by others than individuals high in these qualities. However, two limitations of these studies should be noted. First, they did not explore toxic and socially aversive personalities, which could explain more of the variance in workplace ostracism than the Big Five. Second, they did not examine the mediators and moderators of the relevant effects. Hence, it remains unclear how and when individuals’ personalities lead them to become the target of ostracism. Without knowing the mediators and moderators of ostracism, it is difficult for scholars to identify and recommend effective methods to reduce the occurrence of workplace ostracism. Exploring both mediators and moderators also provides a good opportunity to refine related theories.

In the context of toxic and socially aversive personalities, Dark Triad traits (i.e., Machiavellianism, narcissism, and psychopathy) are considered to have stronger effects on antisocial outcomes than the Big Five. Hence, these traits have attracted the most empirical attention in the deviant behavior literature (e.g., Jonason et al., 2012; Lee & Lim, 2021; Pailing et al., 2014; Paulhus & Williams, 2002; Rauthmann, 2012; Vize et al., 2018; Webster et al., 2016; Webster & Jonason, 2013). The behavior of people with these dark traits in interpersonal situations is often cold, self-serving, and manipulative. People with high levels of these dark personality traits are likely to exhibit behaviors high in agency (“getting ahead”) and low in communion (“getting along”) (Jones & Paulhus, 2011; Rauthmann, 2012); thus, they provoke negative reactions from others. Consequently, people with dark personalities are expected to be socially excluded. Our goal was to develop a theoretical model to investigate whether and how individuals with high levels of Dark Triad traits are exposed to ostracism at work.

The target-behavior mechanism is often used to explain how individual dispositions are related to aggressive behaviors (Nielsen & Knardahl, 2015; Nielsen et al., 2017; Rudert et al., 2020), suggesting that employees with specific dispositions violate expectations, underperform, annoy others, and even go against social norms of polite and friendly interactions, which elicit aggressive behavior from others (Einarsen, 1999; Felson, 1992). According to the target-behavior mechanism, we speculate that targets’ dark personality dispositions may trigger an intentional desire in others (sources) to ostracize them. Furthermore, victim precipitation theory suggests that toxic personality traits predispose individuals to exhibit cognitive, emotional, or behavioral responses that encourage victimization (Bowling et al., 2010; Henle & Gross, 2014). Although this theory has its origins in criminology (Curtis, 1974; Olweus, 1978), a large number of studies have shown that the victim precipitation hypothesis is a good explanation for workplace victimization (Aquino & Bradfield, 2000; Cortina et al., 2018), such as abusive supervision (Henle & Gross, 2014), workplace bullying (Samnani & Singh, 2016), and workplace ostracism (Howard et al., 2020; Wu et al., 2011). Thus, this theoretical foundation not only underlies the relationships between employee personality traits and workplace ostracism, but also offers a cognitive perspective to explain why personality characteristics are associated with exposure to workplace ostracism. We propose that employees with Dark Triad traits elicit workplace ostracism through self-serving cognitions, which refer to individuals’ concern for themselves, or a tunneled, self-interested focus (De Dreu & Nauta, 2009; Mitchell et al., 2018). Individuals with dark personalities usually strive to achieve self-serving goals at the expense of others (Jones & Paulhus, 2010; Rauthmann, 2012); therefore, driven by self-serving cognitions, these people often provoke their own victimization. As these individuals may appear to be “bad apples,” they are likely to be excluded by others at work (Rudert et al., 2020).

Moreover, victim precipitation theory asserts that not all individuals disclose their characteristics in the same way, such that some tend to conceal their characteristics to avoid risks when interacting with others (Curtis, 1974; Olweus, 1978). We posit that people with a strong collectivism orientation prototypically repair the harm caused by their unfavorable characteristics. Therefore, the self-serving cognition mechanism that explains the relationship between Dark Triad traits and workplace ostracism may be moderated by collectivism orientation. Collectivist culture emphasizes interdependence, security, group harmony, good interpersonal relationships, and the norm that group interests take precedence over personal interests (Triandis, 1994, 2001). Triandis (1995) also identified differences among individuals within a single cultural context. Therefore, we focus on collectivism orientation as an individual difference, which refers to the extent to which individuals are willing to sacrifice their personal interests for the benefit of the group, view themselves as interdependent with others, and feel a strong need to fit in the collective or group (Chen & Jing, 2012; Oyserman & Lee, 2008; Sang, 2019; Triandis, 1995). In short, we propose that when individuals have a strong collectivism orientation, their intense collective consciousness maintains their general interest in appearing friendly and socially desirable, thereby mitigating the effect of self-serving cognitions on ostracism.

Our study contributes to the literature on workplace ostracism. First, by revealing whether people with Dark Triad traits are more likely to be ostracized, we enrich victim precipitation theory and extend our knowledge of the antecedents of workplace ostracism, showing that targets’ dark traits increase the likelihood of ostracism. Second, we identify the mediating role of self-serving cognitions, revealing how Dark Triad traits cause workplace ostracism. This finding contributes to the literature on self-serving cognitions, expanding our understanding of the antecedents of workplace ostracism by responding to the call to investigate the mechanisms by which personalities predict workplace ostracism (Mao et al., 2018) and showing that having Dark Triad traits increases the risk of ostracism via self-serving cognitions in the workplace. Third, we examine collectivism orientation as a key moderator to provide a boundary condition for the effect of self-serving cognitions on workplace ostracism, thereby refining victim precipitation theory, which posits that exhibiting certain characteristics does not necessarily result in exclusion. This approach can also help identify who suffers most from self-serving cognitions and provide important practical implications for understanding the conditions under which the influence of self-serving cognitions on workplace ostracism is least pronounced. Our theoretical model is presented in Fig. 1.

Fig. 1
figure 1

Theoretical model

Theory and hypotheses

Dark Triad traits and workplace ostracism

Personality traits represent a form of highly stable dispositions and tendencies to think, feel, and act in certain consistent ways (Pervin, 1993). The target behavior mechanism suggests that employees with specific dispositions violate expectations, underperform, and even break social norms of polite and friendly interactions, triggering aggressive behaviors in others (Einarsen, 1999; Felson, 1992). Similarly, victim precipitation theory suggests that people who exhibit aversive behaviors do not select their targets randomly, as certain characteristics attract such behaviors (Curtis, 1974). Two types of people are likely to be victims (Olweus, 1978). One type is the submissive victim who is anxious, insecure, and quiet, signaling that they are weak and have few defenses against victimization. The second type is the provocative victim who is anxious and aggressive, indicating that they are hostile toward other people. The arguments set forth above explain how Dark Triad traits affect the likelihood of becoming a target of ostracism.

The Dark Triad traits include three socially aversive personalities: Machiavellianism, narcissism, and psychopathy (Furnham et al., 2013; Paulhus & Williams, 2002). Machiavellians appear to be appreciated in certain circumstances (e.g., Deluga, 2001). Narcissists may be popular, charming, and well-liked at first (e.g., Back et al., 2010). Finally, owing to their charming (superficial) behavior and interpersonal manipulation, psychopaths may appear at first like narcissists (i.e., interesting, entertaining) (Rauthmann, 2012). However, further interaction with them shows that they are “toxic” because they tend to act in self-interested and exploitative ways in interpersonal situations (Rauthmann, 2012). People with these dispositions do not conform to social rules and expectations, easily triggering antisocial outcomes, such as bullying (Baughman et al., 2012), violence (Pailing et al., 2014), and aggression (Webster et al., 2016). Ostracism as a punitive response is common in the workplace (e.g., Scott et al., 2013). Below, we explain how each of the Dark Triad traits can lead to individuals being ostracized and present our predictions.

Machiavellianism describes a cold, selfish, manipulative propensity, driven by instrumentality, to act in pragmatic ways, motivated by self-interest, emotionalism, and immoral beliefs (Christie & Geis, 1970; Jones & Paulhus, 2010; Paulhus & Williams, 2002; Rauthmann & Will, 2011). Several studies have shown that Machiavellianism is positively related to hostility (e.g., Jonason & Webster, 2010; Webster et al., 2016). Individuals with a high level of Machiavellianism are characterized by dispositional cynicism and emotional detachment, are manipulative, and satisfy their antisocial desires through political means (Jones, 2013). These characteristics and actions may be repulsive to others, resulting in low relationship satisfaction (Brewer & Abell, 2017) and increasing the risk of being bullied (Baughman et al., 2012) and the likelihood of violence (Pailing et al., 2014) or aggressive responses (Webster et al., 2016). Because Machiavellianism is linked to these reactive antisocial outcomes, we contend that there is a strong link between Machiavellianism and being ostracized. Based on the target behavior mechanism, the tendency to behave in a Machiavellian manner breaks social norms and easily elicits negative reactions from others in the interpersonal circumplex (Jones & Paulhus, 2010, 2011; Paulhus & Williams, 2002; Rauthmann, 2012). Individuals high in Machiavellianism are likely to be at high risk of becoming a target of ostracism. Thus, we propose the following hypothesis:

Hypothesis 1

Machiavellianism is positively related to workplace ostracism.

Narcissism is characterized by grandiosity, entitlement, dominance, and superiority (Paulhus & Williams, 2002), describing individuals’ tendency to hold grandiose and inflated self-views, exhibit extreme vanity, seek attention and admiration, and have feelings of superiority, authority, and entitlement (Raskin & Terry, 1988; Rauthmann, 2012). Research has shown that narcissism is linked to antisocial motivation (Jones & Paulhus, 2010; Paulhus & Williams, 2002); thus, narcissists are likely to exhibit interpersonal aggression (Webster et al., 2016). According to the target-behavior mechanism, narcissists display coldness and dominance (Edershile & Wright, 2021), violate the norms of interpersonal interactions, and are perceived with dislike and social aversion, which become less and less favorable as interactions and relationships progress (Back et al., 2010; Edershile & Wright, 2021; Rauthmann, 2012). Such unfavorable situations easily trigger punitive deviant behavior toward narcissists (Pailing et al., 2014). Because ostracism, as a punishment, is likely to be a reaction to highly narcissistic individuals, we propose the following hypothesis:

Hypothesis 2

Narcissism is positively related to workplace ostracism.

Psychopathy is characterized by callousness, thrill-seeking, impulsivity, and a lack of empathy; it is considered the most toxic characteristic of the Dark Triad because of its links to extreme and frequent antisocial behavior (Patrick, 2007; Paulhus & Williams, 2002; Vize et al., 2018). According to the target-behavior mechanism, individuals high in psychopathy are dispositional rule-breakers (Hare & Neumann, 2005), have low regard for others, and are characterized by disagreeableness and impulsiveness, all of which make them repulsive to others (Hare, 2003; Rauthmann, 2012). Many studies have shown that psychopathy is linked to anxiety and that people with psychopathy have an increased propensity for aggression and violence (Falkenbach et al., 2021; Pailing et al., 2014). Therefore, we posit that individuals high in psychopathy tend to push others to exclude or ignore them. Thus, we propose the following hypothesis:

Hypothesis 3

Psychopathy is positively related to workplace ostracism.

The mediating role of self-serving cognitions

Self-serving cognitions are a cognitive state of self-interest characterized by attempts to obtain personal benefit (De Dreu & Nauta, 2009; Mitchell et al., 2018). Because individuals with self-serving cognitions are concerned about their own needs, benefits, and interests, they concentrate on enhancing their work benefits and protecting themselves at work. These tendencies violate the principle of social exchange, which emphasizes reciprocity, resource exchange, and exchange relationships (Blau, 1964; Cropanzano & Mitchell, 2005).

The Dark Triad traits consist of three distinct but overlapping dimensions (Jones & Paulhus, 2011; Pailing et al., 2014; Paulhus & Williams, 2002; Rauthmann, 2012). People with these dark personalities are typically cold and self-interested; they tend to engage in manipulative behaviors; and they are usually toxic and antisocial in interpersonal situations (Paulhus & Williams, 2002; Rauthmann, 2012). Davis (1990) argued that individuals with antisocial personality disorders hold many self-serving beliefs. Machiavellians are selfish and self-centered and tend to manipulate others to pursue selfish, agentic goals (Christie & Geis, 1970; Rauthmann & Will, 2011). In their self-interested worldview, such people tend to care only about themselves and strive for self-interested goals, which translates into self-serving cognitions. Narcissists tend to exhibit extreme vanity, seek attention and admiration from others, and have feelings of superiority. In addition, their inflated self-views are acquired by devaluing others (Rauthmann, 2012). This tendency reflects a focus on self-interest, which is a type of self-serving cognitions (Mitchell et al., 2018). Psychopaths have cold affect, low empathy, and low regard for others and tend to impulsively seek thrills (Paulhus & Williams, 2002; Rauthmann, 2012), suggesting negative other-models, which are often accompanied by positive self-models (Rauthmann, 2012). Taken together, Dark Triad traits are toxic and antisocial; therefore, employees with these traits are self-centered and self-oriented, strive to pursue self-interest, and easily develop self-serving cognitions in the workplace. Hence, we propose the following hypothesis:

Hypothesis 4

(a) Machiavellianism, (b) narcissism, and (c) psychopathy are positively related to self-serving cognitions.

Self-serving cognitions can cause employees to become the target of ostracism. Employees driven by self-serving cognitions tend to engage in self-interested efforts and focus on taking advantage of any opportunity to gain personal advantage, even if their actions harm others (Nagin et al., 2002). These heightened self-serving cognitions may have negative consequences, such as an increased tendency for cheating (Mitchell et al., 2018) and antisocial behavior (Van Leeuwen et al., 2014).

As discussed above, victim precipitation theory suggests that perpetrators choose their targets based on certain characteristics (Curtis, 1974). It is uncomfortable to ostracize others because ostracism violates norms of inclusion and may trigger retaliatory behaviors (Li et al., 2021). Hence, perpetrators usually have good reason to engage in ostracism and believe that other people will recognize their ostracism. Consequently, ostracism occurs when individuals are viewed as poor exchange partners who violate group or social norms (Scott et al., 2013).

Employees with self-serving cognitions tend to manifest these cognitions through self-interested behaviors; they focus on their own interests and lack motivation to seek mutual benefits by acting uncooperatively and ignoring the norms established and maintained by their work group (Mitchell et al., 2018). This selfishness violates the principle of mutually beneficial social exchange relationships and may put the work group at risk of deviating from the group norm. As a result, employees with a high degree of self-serving cognitions appear to their colleagues as bad coworkers and poor exchange partners. One way to punish them is to ostracize them, forcing them to suppress their undesirable characteristics and modify their unfavorable behavior (Wu et al., 2021). Therefore, we suggest that employees with self-serving cognitions may cause workplace ostracism and propose the following hypothesis accordingly:

Hypothesis 5

Employees’ self-serving cognitions are positively related to workplace ostracism.

Our arguments set forth above show that employees with Dark Triad traits, characterized by cold, self-interested, and antisocial behavior, tend to focus on self-interest, thereby forming self-serving cognitions. These individuals’ self-serving cognitions violate reciprocal norms, making them poor exchange colleagues and, consequently, the likely targets of ostracism. In summary, we expect employees with Dark Triad traits to experience workplace ostracism because of their self-serving cognitions. Thus, we propose the following hypothesis:

Hypothesis 6

Employees’ self-serving cognitions mediate the positive relationships between (a) Machiavellianism, (b) narcissism, and (c) psychopathy and workplace ostracism.

The moderating role of collectivism orientation

Collectivism orientation refers to the extent to which an individual views the self as an aspect of a collective or in-group, emphasizing the importance of the interpersonal self, interdependent relationships, and reciprocal relationships (Brewer & Chen, 2007; Hofstede, 1984, 2001). A meta-analysis conducted by Oyserman et al. (2002) demonstrated that collectivism orientation had a significant influence on basic psychological domains, including self-concept, well-being, attributional style, and relationality.

Based on victim precipitation theory (Curtis, 1974), we posit that collectivism orientation as a contingent factor moderates the relationship between self-serving cognitions and workplace ostracism. Victim precipitation theory (Curtis, 1974) argues that not all individuals display undesirable characteristics that may cause aversive behavior. Some individuals may choose to conceal these characteristics to reduce the likelihood of exclusion. We argue that employees with a strong collectivism orientation are such individuals. Brewer and Chen (2007) pointed out that collectivism orientation has a profound influence on self-concept, causing individuals to view themselves as part of an in-group and to value interpersonal interdependence. Employees with a strong collectivism orientation prioritize the interests of the group and value group harmony and solidarity (Triandis, 2001). They tend to care about others and maintain good interpersonal relationships, such as helping coworkers, which means that they come across as friendly and good coworkers. These positive self-descriptions present a good image, which helps to reduce their selfish performance and increase their social desirability (Steenkamp et al., 2010). Thus, employees with a strong collectivism orientation pay considerable attention to preventing the spiraling effect of self-serving cognitions on the violation of social or group norms. They leverage their social adaptivity and attempt to establish and maintain good relationships with others. These behavioral tendencies help to compensate for the detrimental effect of self-serving cognitions in the workplace (i.e., ostracism). Accordingly, collectivism orientation weakens the relationship between self-serving cognitions and workplace ostracism. In contrast, employees with a weak collectivism orientation care little about others and harm interpersonal interdependence and exchange relationships, which is consistent with self-serving cognitions, amplifying their image as poor exchange coworkers. Therefore, we predict that employees with a weak collectivism orientation may face more negative reactions to their self-serving cognitions; that is, they may be more likely to be ostracized in the workplace. Therefore, we propose the following hypothesis:

Hypothesis 7

Collectivism orientation moderates the positive relationship between self-serving cognitions and workplace ostracism, such that this relationship is weaker when employees’ collectivism orientation is high rather than low.

According to the moderated mediation logic of Edwards and Lambert (2007), we expect to find an integrated framework in which self-serving cognitions mediate the positive relationship between Dark Triad traits and workplace ostracism, and collectivism orientation moderates the relationship between self-serving cognitions and workplace ostracism. Based on the idea that Dark Triad traits are positively related to self-serving cognitions and that the relationship between self-serving cognitions and workplace ostracism is weaker when collectivism orientation is high rather than low, we expect collectivism orientation to moderate the strength of the mediating effect of self-serving cognitions on the relationship between Dark Triad traits and workplace ostracism. Because we predict a weaker relationship between self-serving cognitions and workplace ostracism when collectivism orientation is at high rather than low levels, we expect the positive indirect effect of the Dark Triad traits on workplace ostracism through self-serving cognitions to be weaker among employees with a strong collectivism orientation than among those with a weak collectivism orientation. Specifically, employees’ Dark Triad traits lead to self-serving cognitions, but these cognitions are less likely to make them targets of ostracism when they have a high level of collectivism orientation than when they have a low level of collectivism orientation.

Hypothesis 8

Collectivism orientation moderates the indirect effect of self-serving cognitions on the relationships between (a) Machiavellianism, (b) narcissism, and (c) psychopathy and workplace ostracism, such that the indirect effect is weaker when employees’ collectivism orientation is high rather than low.

Overview of the study

We conducted two studies: an experimental study (Study 1) and a multi-wave field survey study (Study 2). In Study 1, to address potential reverse causality and respond to the call of Wu et al. (2015), we first experimentally tested the causal effect of the Dark Triad traits on ostracism intentions (Hypotheses 1–3). In Study 2, we applied a multi-wave design to test our full model (Hypotheses 1–8) in a field setting.

Study 1 method

Participants

The participants were recruited online. They did not come to the laboratory in person but were sent a link via WeChat (a multipurpose messaging application with over 1 billion active users; Qin et al., 2020) to participate in the online study (the research materials were designed and implemented on the SurveyStar platform). We ultimately recruited 328 adults (134 men, 194 women, Mage = 22.84 years, SD = 7.36) in China. Regarding their education level, 35 of the participants (10.67%) had completed high school or less, 203 (61.89%) held an associate degree, 58 (17.68%) held a bachelor’s degree, 31 (9.45%) held a master’s degree or above, and one person did not provide their education level.

Materials and procedure

The materials and procedure for this experiment followed Rudert et al. (2020). The experiment involved a between-subjects design; 328 participants were randomly assigned to one of seven conditions (Machiavellianism: high vs. low; narcissism: high vs. low; psychopathy: high vs. low; and a control condition). The participants read a vignette describing a person named Xiao Ming. The basic version of the vignette contained no information about Xiao Ming’s personality. The vignette was as follows (Rudert et al., 2020):

Xiao Ming is 20 years old and works part-time at a nearby restaurant. In his free time, he likes watching movies, listening to music, and spending time outdoors. On a typical day, Xiao Ming goes to class and spends time on his computer. After dinner, he usually watches TV. His favorite shows are crime series and he likes quiz shows.

The participants in the control condition received the basic version of the vignette, while the vignettes for the remaining six conditions were modified so that Xiao Ming was also described as being either high or low in one of the manipulated Dark Triad traits (i.e., Machiavellianism, narcissism, or psychopathy). Table 1 provides details.

Table 1 The descriptions of the manipulated dark traits dimension

After reading the vignette, the participants first completed an attention question (“Xiao Ming hates crime series/likes crime series”; 7-point scale) and a manipulation check on the personality dimension manipulated in their respective conditions (e.g., “Xiao Ming is low in Machiavellianism/high in Machiavellianism”; 7-point scale). In the control condition, the participants rated Xiao Ming on all the Dark Triad traits. Next, the participants were asked to imagine that Xiao Ming wanted to join a club to which they belonged and to rate their intention to ostracize Xiao Ming. Cronbach’s α was 0.87, with one of the seven items being “I might find myself ignoring Xiao Ming” (Hales, Kassner, Williams, & Graziano, 2016). Finally, after providing demographic information, the participants were paid in cash and thanked for their participation.

Study 1 results

Manipulation checks

We tested the success of our manipulation using three one-way analyses of variance (ANOVAs) with the manipulated dark trait dimensions (high vs. control vs. low) as independent variables and the corresponding trait questions as dependent variables. The results of the ANOVAs showed significant effects of the manipulated traits: Machiavellianism (F = (2, 131) = 6.94, p < 0.01, η2 = 0.10), narcissism (F = (2, 134) = 25.46, p < 0.001, η2 = 0.28), and psychopathy (F = (2, 148) = 18.38, p < 0.001, η2 = 0.20). Least significant difference (LSD) post hoc analyses showed that the effect of the high conditions was significantly greater than that of the low or control conditions and that the effect of the low conditions was not significantly different from that of the control condition for Machiavellianism, narcissism, and psychopathy. All descriptive statistics are presented in Table 2.

Table 2 Means (and Standard Deviations) of the manipulation checks as a function of the manipulated dark traits dimension in Study 1

Dependent variables

The three one-way ANOVAs (manipulated Dark Triad dimension: high vs. control vs. low) on the participants’ intention to ostracize Xiao Ming revealed significant differences for Machiavellianism (F = (2, 131) = 4.274, p < 0.05, η2 = 0.06), narcissism (F = (2, 134) = 7.02, p < 0.01, η2 = 0.09), and psychopathy (F = (2, 148) = 9.66, p < 0.001, η2 = 0.12). LSD post hoc analyses showed that the participants tended to have a stronger intention to ostracize a target high in Machiavellianism, narcissism, and psychopathy than a target in the control group. Additionally, the participants tended to have a stronger intention to ostracize a target high in narcissism and psychopathy than a target low in narcissism and psychopathy. Overall, Hypotheses 1–3 were supported. All descriptive statistics are reported in Table 3 and Fig. 2.

Table 3 Means (and Standard Deviations) of participants’ intention to ostracize Xiao Ming as a function of the manipulated dark traits dimension in Study 1
Fig. 2
figure 2

Mean ostracism intentions as a function of the manipulated dark traits dimension in Study 1

Study 2 method

Sample and procedure

Because ostracism is a universal phenomenon that occurs in all workplaces (Ferris et al., 2008), we collected data from various types of organizations in China, including government departments and public institutions, state-owned enterprises, private enterprises, and foreign enterprises/joint ventures. We first designed the questionnaire through an online platform (“SurveyStar,” Changsha Ranxing IT Ltd.) and set the requirement to complete all items. Then, two researchers adopted a snowball sampling strategy on WeChat to invite employees to participate in our survey. After collecting two waves of data with a one-month interval, we obtained a final sample of 239 participants. In the first wave, 462 participants were invited to participate in the survey and asked to provide information on their demographic characteristics (gender, age, education level, position, organizational tenure, organization type), Machiavellianism, narcissism, psychopathy, self-serving cognitions, and collectivism orientation. One month after the first wave, the participants were invited to participate in our survey, asking them to report their perceptions of workplace ostracism over the last month. Ultimately, 239 participants responded, giving a response rate of 51.73%.

Because the attrition rate of the sample was relatively high, we compared systematic differences between a group of participants who participated only at T1 (a total of 462 participants) and a group who participated at both T1 and T2 (a total of 239 participants). The results indicated that the significance of Levene’ test for equality of variance for all the variables (e.g., Machiavellianism, p = 0.16 > 0.05) were all greater than 0.05, indicating that the difference between the two groups was not statistically significant.

Of the 239 employees, 109 were men and 130 were women. The average age of the participants was 29.25 years (SD = 6.24) and their average organizational tenure was 3.63 years (SD = 5.29). In terms of education level, 127 participants (53.14%) held a bachelor’s degree, 67 (28.03%) held a master’s degree or above, 25 (10.46%) held an associate degree, and 20 (8.37%) had completed high school or less. Most of the participants (176, 73.64%) were frontline employees, 43 (17.99%) were frontline managers, 14 (5.86%) were mid-level managers, and six (2.51%) were senior managers. The participants worked in a variety of organizations, including government departments and public institutions (106 participants, 44.35%), state-owned enterprises (43 participants, 17.99%), private enterprises (77 participants, 32.22%), and foreign enterprises/joint ventures (13 participants, 5.44%). All categorical variables (i.e., gender, education level, position, organization type) were transformed into dummy variables and then used in related analyses.

Measures

Because the measure of self-serving cognitions was originally developed in English, we translated it into Chinese following Brislin’s (1986) back-translation procedure. Chinese versions were available for the other three variables (i.e., Dark Triad traits, collectivism orientation, and workplace ostracism) and had previously been applied and validated in the Chinese setting. The participants rated these measures using a 5-point Likert-type scale, ranging from 1 (“strongly disagree”) to 5 (“strongly agree”).

The Dark Triad. We measured Machiavellianism, narcissism, and psychopathy using the 12-item scale developed by Jonason and Webster (2010); this measure was validated in the Chinese cultural context by Geng et al. (2015). Each of the three sub-scales consisted of four items, with sample items including “I tend to manipulate others to get my way” (Machiavellianism: Cronbach’s α = 0.83, McDonald’s ω = 0.84), “I tend to want others to admire me” (narcissism: Cronbach’s α = 0.84, McDonald’s ω = 0.85), and “I tend to be unconcerned with the morality of my actions” (psychopathy: Cronbach’s α = 0.71, McDonald’s ω = 0.74).

Self-serving cognitions. We measured self-serving cognitions using the four items from Mitchell et al. (2018), adapted from De Dreu and Nauta’s (2009) measure of self-concern. A sample item is “I am concerned about my own needs and interests” (Cronbach’s α = 0.76, McDonald’s ω = 0.78).

Collectivism orientation. We measured collectivism orientation using the six-item scale originally developed by Dorfman and Howell (1988) and later applied in the Chinese setting by Xu et al. (2020). A sample item is “Group welfare is more important than individual rewards” (Cronbach’s α = 0.89, McDonald’s ω = 0.89).

Workplace ostracism. We measured workplace ostracism using the 10-item scale originally developed by Ferris et al. (2008) and later applied in the Chinese setting by Wu et al. (2012). A sample item is “I noticed others would not look at me at work” (Cronbach’s α = 0.95, McDonald’s ω = 0.95).

Control variables. We controlled for the participants’ demographic characteristics, including gender (1 = male, 2 = female), age, education level (1 = high school or less, 2 = associate degree, 3 = bachelor’s degree, 4 = master’s degree or above), position (1 = frontline employee, 2 = frontline manager, 3 = mid-level manager, 4 = senior manager), organizational tenure, and organization type (1 = state-owned enterprise, 2 = private enterprise, 3 = foreign enterprise/joint venture, 4 = government department and public institution).

Study 2 results

Confirmatory factor analyses

We conducted confirmatory factor analyses (CFAs) using Mplus 8.0 to evaluate the distinctiveness of the six variables (i.e., Machiavellianism, narcissism, psychopathy, self-serving cognitions, collectivism orientation, and workplace ostracism). Due to the relatively small sample size, we first parceled collectivism orientation and workplace ostracism into three and five items, respectively (Little et al., 2002). The CFA results are presented in Table 4, showing that the six-factor model had a better fit (χ2(df) = 564.71 (237), p < 0.001, CFI = 0.91, TLI = 0.89, RMSEA = 0.076, SRMR = 0.072) than the alternative models. Therefore, the discriminant validity of the six key measures was confirmed. We also conducted a post hoc power analysis of the model fit test. The results showed that its power was 0.203 and its matching effect was preferable.

Table 4 Result of Confirmatory Factor Analyses

Furthermore, following Podsakoff et al. (2003), we examined the change in the fit indices of the seven-factor model when adding a common latent factor. If the fit indices of the model increased significantly after adding the common latent factor (i.e., if CFI and TLI increased by more than 0.1 and RMSEA and SRMR decreased by more than 0.05), this would indicate severe common method bias. As shown in Table 4, there was no significant change in the fit indices of the seven-factor model when adding the common latent factor (χ2(df) = 541.32 (229), p < 0.001, CFI = 0.91, TLI = 0.90, RMSEA = 0.076, SRMR = 0.075). This showed that common method bias was not a major problem in our study.

Descriptive statistics and correlations

The means, standard deviations, and correlations of the variables are presented in Table 5. The results showed that Machiavellianism (r = 0.36, p < 0.001) and narcissism (r = 0.30, p < 0.001) were correlated with workplace ostracism. Machiavellianism (r = 0.37, p < 0.001), narcissism (r = 0.35, p < 0.001), and psychopathy (r = 0.28, p < 0.001) were positively correlated with self-serving cognitions, and self-serving cognitions were positively correlated with workplace ostracism (r = 0.27, p < 0.001).

Table 5 Means, standard deviations and correlations among variables

Hypothesis testing

We conducted hierarchical regression analysis with SPSS 22.0 and bootstrap analyses with Hayes’ (2013) PROCESS macro and Mplus 8.0 to test our hypotheses.

Main effects. The regression results are shown in Table 6. The results indicated that Machiavellianism (B = 0.13, n.s.) and narcissism (B = 0.08, n.s.) were not significantly related to workplace ostracism, while psychopathy (B = 0.15, p < 0.05) was positively and significantly related to workplace ostracism, thus supporting Hypothesis 3 but not Hypotheses 1 and 2. In addition, Machiavellianism (B = 0.20, p < 0.01) and narcissism (B = 0.20, p < 0.01) were positively and significantly related to self-serving cognitions. Psychopathy was not significantly related to self-serving cognitions (B = 0.07, n.s.). However, the bootstrap analysis results showed that psychopathy was positively and significantly related to self-serving cognitions (B = 0.27, SE = 0.06, 95% confidence interval [CI] = [0.15, 0.39]). Self-serving cognitions were positively and significantly related to workplace ostracism (B = 0.21, p < 0.01). Thus, these results supported Hypotheses 4 and 5.

Table 6 Results of multiple regression analyses

Mediating effect of self-serving cognitions. The bootstrap analyses revealed that Machiavellianism, narcissism, and psychopathy had indirect effects on workplace ostracism via self-serving cognitions. The indirect effects of Machiavellianism (B = 0.06, SE = 0.03, 95% CI = [0.004, 0.12]), narcissism (B = 0.06, SE = 0.03, 95% CI = [0.01, 0.13]), and psychopathy (B = 0.08, SE = 0.03, 95% CI = [0.03, 0.15]) on workplace ostracism via self-serving cognitions were positive and significant, thus supporting Hypothesis 6.

Moderating effect of collectivism orientation. We performed centering before examining the moderating effect of collectivism orientation because centering can reduce nonessential ill-conditioning (e.g., multicollinearity). We centered self-serving cognitions and collectivism orientation and calculated a new interaction term using the centered variables, then conducted the hierarchical moderated regression analyses. The regression results for the moderating effect indicated that the interaction term between self-serving cognitions and collectivism orientation was negatively and significantly related to workplace ostracism (B = − 0.27, p < 0.01). Figure 3 shows that the positive relationship between self-serving cognitions and workplace ostracism was significant when collectivism orientation was low (B = 0.37, p < 0.001) rather than high (B = 0.05, n.s.), thus supporting Hypothesis 7.

Fig. 3
figure 3

The moderating effect of collectivism orientation

Finally, we tested moderated mediation using Mplus 8.0. The results shown in Table 7 indicated that the indirect effect of Machiavellianism on workplace ostracism via self-serving cognitions was stronger when collectivism orientation was low (β = − 0.10, SE = 0.05, 95% CI = [0.02, 0.19]) than when it was high (β = − 0.04, SE = 0.04, 95% CI = [–0.04, 0.10]), but the difference was not significant (β = − 0.06, SE = 0.05, 95% CI = [–0.16, 0.04]). Similarly, the indirect effect of narcissism on workplace ostracism via self-serving cognitions was stronger when collectivism orientation was low (β = 0.12, SE = 0.05, 95% CI = [0.04, 0.21]) than when it was high (β = 0.06, SE = 0.04, 95% CI = [–0.004, 0.14]), but the difference was not significant (β = − 0.05, SE = 0.05, 95% CI = [–0.15, 0.04]). The indirect effect of psychopathy on workplace ostracism via self-serving cognitions was stronger when collectivism orientation was low (β = 0.10, SE = 0.04, 95% CI = [0.03, 0.20]) than when it was high (β = 0.03, SE = 0.03, 95% CI = [–0.03, 0.09]), but the difference was not significant (β = − 0.07, SE = 0.05, 95% CI = [–0.17, 0.01]). Thus, these results did not support Hypothesis 8.

Table 7 Results of moderated mediation

General discussion

Ostracism can be caused by a variety of factors, including the personality disposition of the ostracized target (Rudert et al., 2020). Our research examined whether Dark Triad traits can elicit ostracism. In the experimental study, we found that individuals high in Machiavellianism, narcissism, and psychopathy were likely to be ostracized by others. However, our field study found that Machiavellianism and narcissism were not significantly related to workplace ostracism. This finding may be the result of the two faces of Machiavellianism and narcissism. Indeed, there are mixed findings regarding how Machiavellians and narcissists are perceived by others. For instance, Machiavellians seem well-liked in certain circumstances (Deluga, 2001) but are perceived more negatively after prolonged interactions (Wilson et al., 1998). Narcissists may be popular, charming, and well-liked at first, but later their likability decreases (Back et al., 2010; Wink, 1991). These mixed impressions may lead Machiavellianism and narcissism to have nonsignificant effects on workplace ostracism.

Based on victim precipitation theory (Curtis, 1974; Olweus, 1978), we further examined the mediating role of self-serving cognitions and the moderating role of collectivism orientation. Machiavellianism, narcissism, and psychopathy showed significant relationships with self-serving cognitions, causing exposure to workplace ostracism. When the participants had a strong collectivism orientation, the effect of self-serving cognitions on workplace ostracism was not significant. In contrast, when they had a weak collectivism orientation, self-serving cognitions led to workplace ostracism.

Theoretical implications

This study makes several important theoretical contributions. First, our research contributes to the literature on the relationships between personality dispositions and workplace ostracism. Prior studies have shown that the Big Five are associated with workplace ostracism (Rudert et al., 2020; Wu et al., 2011), that dark traits are weakly correlated or uncorrelated with the Big Five (Vize et al., 2018), and that Dark Triad traits are considered more toxic and more closely related to antisocial outcomes than the Big Five (Jonason et al., 2012; Lee & Lim, 2021; Paulhus & Williams, 2002; Rauthmann, 2012; Webster et al., 2016). These dark traits have been overlooked in research on workplace ostracism. Using the target-behavior mechanism and victim precipitation theory, we investigated the relationship between the Dark Triad traits and workplace ostracism, and our results showed that individuals’ Machiavellianism, narcissism, and psychopathy could lead them to be ostracized at work. Our research therefore shows that Dark Triad traits explain why certain individuals are targets of ostracism, making a significant contribution to the literature on the relationships between personality characteristics and workplace ostracism.

Second, our research contributes to the literature on self-serving cognitions by revealing how the Dark Triad causes workplace ostracism. Using victim precipitation theory (Curtis, 1974; Olweus, 1978), we identified the self-interested processes (i.e., self-serving cognitions) that explain why Dark Triad traits increase the risk of being ostracized. Dodge et al. (2006) suggested that self-serving cognitions mediate the effect of personality traits on social situational behavior. Several studies have explored the mediating role of self-serving cognitions on predictors of antisocial behavior (e.g., Barriga et al., 2001; Mitchell et al., 2018; Van Leeuwen et al., 2014). For instance, Van Leeuwen et al. (2014) found that self-serving cognitions mediated the relationship between callous-unemotional traits and antisocial behavior. Our results revealed that Dark Triad traits tend to create self-serving cognitions, which lead to workplace ostracism. Thus, our findings echo those of Dodge et al. (2006) and extend the literature on self-serving cognitions. In addition, our findings enrich our understanding of the antecedents of workplace ostracism by showing that the Dark Triad can increase the risk of being ostracized via self-serving cognitions in the workplace, which contributes to the literature on workplace ostracism and responds to the call to examine the mechanism by which antecedents predict workplace ostracism (Mao et al., 2018).

Third, our study contributes to research on collectivism orientation by examining collectivism orientation as a key moderator of the boundary condition of the effect of self-serving cognitions on workplace ostracism. The literature has predominantly focused on the moderating role of collectivism orientation in affecting how people respond to ostracism (Mao et al., 2018; Xu et al., 2020). Individuals with a strong collectivism orientation are likely to be sensitive to ostracism and various negative outcomes. We found that individuals with a strong collectivism orientation could counteract the negative effect of self-serving cognitions, thereby reducing the likelihood of ostracism. Hence, our findings provide important insight into the fact that not all people demonstrate cognitions in the same way, thereby contributing to victim precipitation theory (Curtis, 1974; Olweus, 1978) by identifying a boundary condition for this theory. However, our results showed that collectivism orientation did not moderate the indirect effects of Dark Triad traits on workplace ostracism via self-serving cognitions. This finding indicates that other mediators not considered in our research (e.g., mediators that may help to explain how workplace ostracism occurs, such as emotional labor) may be moderated by collectivism orientation. Further research could consider other mediators to explain the mechanism linking Dark Triad traits and workplace ostracism. Overall, our findings extend the literature on collectivism orientation by shedding light on its moderating role in the context of workplace ostracism.

Practical implications

As our results showed that ostracism intentions and behaviors increase primarily because of individuals’ negative personality dispositions (i.e., Machiavellianism, narcissism, and psychopathy), we recommend that greater attention be paid to preventing and reducing workplace ostracism. First, our results showed that the Dark Triad traits were positively associated with ostracism, which means that individuals with such personality dispositions may be at higher risk of ostracism. Accordingly, it would be useful for organizations to conduct personality tests at the time of recruitment to reduce the number of employees with extremely negative personality dispositions. Facing employees with dark personality dispositions, organizations should also adopt interventions aimed at reducing the risk of ostracism. For instance, they could provide training to improve employees’ social skills, because only employees who learn to get along with others in the workplace are successful.

Second, we found that dark employees high in self-serving cognitions were more likely to be ostracized by others. Individuals driven by a cognitive state of self-interest tend to gain advantages at the expense of others and will therefore inevitably experience counterattacks, increasing the likelihood of being ostracized. The harm caused by self-serving cognitions needs to be repaired. Our results showed that employees with a strong collectivism orientation could offset the negative effect of self-serving cognitions caused by the Dark Triad, thus reducing ostracism. The implication of this result is that organizations should pay more attention to cultivating a collectivism orientation to foster intragroup inclusiveness, including caring, harmony, cooperation, selflessness, and collective interests over individual interests. In the process of shaping organizational culture, organizations should emphasize the importance of “we” consciousness, emotional dependence, duties, and the need for stable and predetermined friendship.

Limitations and future directions

We conducted two studies (an experimental study [Study 1] and a multi-wave field study [Study 2]), our study has limitations that should be addressed in future research. First, we collected survey data of Study 2 through self-report, which may have led to common method variance (CMV) (Podsakoff et al., 2003; Podsakoff et al., 2012). To control for common method bias, we implemented procedural and statistical remedies. First, we introduced a temporal separation (i.e., a one-month interval) between our measurement of the predictor and criterion variables. Second, we added a common latent factor, which resulted in no significant change in the fit indices of the seven-factor model with the common latent factor (χ2(df) = 541.32 (229), p < 0.001, CFI = 0.91, TLI = 0.90, RMSEA = 0.076, SRMR = 0.075). Thus, CMV was not a serious issue. Nevertheless, future research should pay attention to CMV and attempt to control for common method bias, such as by measuring the predictor and criterion variables from different sources. In addition, in terms of theory, we focused on the victim and the focal employee’s subjective self-perception, which may have led to bias.

Second, self-reported data of Study 2 raise another concern about perception issues. It is possible that people with high Dark Triad traits and self-serving cognitions may evaluate their experience of being ostracized very differently from people who are low in these characteristics, such that they tend to form significantly higher perceptions of being ostracized by others via a victim mindset. However, there is no obvious evidence that respondents under-report their feelings of exposure to deviant behavior (e.g., workplace ostracism), and thus, self-report data are also appropriate for assessing workplace deviance (Aquino & Douglas, 2003; Yang & Treadway, 2018). Nevertheless, future research could collect data from perpetrators and test our proposed model with these data. As Howard et al. (2020) suggested, focusing on perpetrators could explain the antecedents of felt ostracism. Therefore, it is necessary to shift focus towards on the perception of the perpetrator, or on the perspective between the victim and the perpetrator, or on the triadic social process between the victim, the perpetrator, and the environment, to better understand the characteristics, cognitive, and environmental factors that lead to workplace ostracism.

Third, our studies were conducted in the Chinese context, which raises concerns about the generalizability of our results. Chinese cultures emphasize the importance of collectivism; thus, our participants were more likely to be affected by a collectivism orientation than non-Chinese participants would have been. Therefore, it would be useful to conduct cross-cultural research (in both Chinese and Western contexts) to capture a wider range of collectivism orientation in the future.

Fourth, Dark Triad traits and self-serving cognitions were measured at Time 1, and their causal relationship could not be determined. Accordingly, future research should conduct a three-wave survey (to measure the independent, mediating, and dependent variables) or a longitudinal study. Our research primarily focused on the mediating role of self-serving cognitions, and our results showed that collectivism orientation did not moderate the indirect effect. However, the possible influence of other mediators should be considered. Therefore, future research should examine other potential mediators, such as emotional labor, to explain the relationship between Dark Triad traits and workplace ostracism (Brotheridge & Lee, 2003; Grandey & Melloy, 2017). In addition, future research should consider the characteristics of other people (e.g., bystanders, sources) to investigate their reactions to the self-serving cognitions of focal employees, which would further improve our understanding of how ostracism is generated.

Fifth, although personality traits influence whether a person becomes a target of ostracism, the work context is also important. We focused on personality traits and did not address situational factors that may trigger workplace ostracism. Future research should pay attention to specific situations (e.g., social climate) to extend the literature on the antecedents of workplace ostracism.

Conclusion

Drawing on the target-behavior mechanism and victim precipitation theory, we found that employees with Dark Triad traits (i.e., Machiavellianism, narcissism, and psychopathy) are at higher risk of becoming the target of ostracism in the workplace. Self-serving cognitions explain why employees high in Dark Triad traits are ostracized, while collectivism orientation weakens the positive relationship between self-serving cognitions and workplace ostracism. These findings enhance our understanding of why and when ostracism occurs. We hope that our study will encourage further research into workplace ostracism.