Ho Kwong Kwan, Department of Management, Academic Building, Drexel University, 101 North 33rd Street, Philadelphia, PA 19104, USA (firstname.lastname@example.org).
The study reported here examined the relationship between workplace ostracism and employee psychological distress (i.e. job tension, emotional exhaustion, and depressed mood at work) by focusing on the joint moderating effects of ingratiation and political skill. Data from a two-wave survey of 215 employees in two oil and gas firms in China indicated that as predicted, workplace ostracism was positively related to psychological distress. Moreover, the findings showed that when employee political skill was high, ingratiation neutralized the relationship between workplace ostracism and psychological distress, but when it was low, ingratiation exacerbated the relationship.
Workplace ostracism, which is the extent to which individuals perceive that they are ignored or excluded by other employees in the workplace, is a pervasive workplace phenomenon (Ferris et al., 2008). A survey of 262 full-time employees indicated that over five years, 66 per cent of respondents felt that they were systematically ignored by colleagues, while 29 per cent reported that other people intentionally left the area when they entered (Fox and Stallworth, 2005). Workplace ostracism decreases the opportunity for social interaction, which is essential in enabling people to fulfil their psychological needs. Indeed, workplace ostracism potentially influences employees' mental and physical health (Heaphy and Dutton, 2008). This is particularly the case today as teamwork has increased dramatically, implying the need for more frequent social interaction and communication with colleagues (Sundstrom et al., 2000). A recent study has shown that workplace ostracism is a very influential variable in explaining a thwarted sense of belonging and reduced workplace contributions (O'Reilly and Robinson, 2009). Despite the prevalence and importance of workplace ostracism, surprisingly little research has examined the impact of this phenomenon (Ferris et al., 2008). Thus, it is both crucial and timely to understand the effect of being ostracized on employee work outcomes.
Ostracism is an interpersonal stressor that can lead to psychological distress (Williams, 1997, 2001). Research has shown that distress experienced in the workplace is strongly correlated with undesirable outcomes such as life distress, turnover intention, and poor physical health (Grandey and Cropanzano, 1999). Consequently, it is important to study the relationship between workplace ostracism and psychological distress.
Moreover, understanding how to cope with ostracism is also critical because effective coping strategies may mitigate the relationships between ostracism and its negative outcomes (Williams, 2007). A common behavioural strategy for coping with ostracism is that of ingratiation (Williams and Zadro, 2005). As a type of social influence behaviour, ingratiation is ‘an attempt by individuals to increase their attractiveness in the eyes of others’ (Liden and Mitchell, 1988, p. 572). Examples include conformity, flattery, and doing favours for others (Jones and Pittman, 1982). This study focuses on ingratiation, because it is a powerful tactic commonly applied at an initial stage in interpersonal interactions across many contexts (Treadway et al., 2007). However, research suggests that the success of behavioural strategies tends to vary between individuals (Williams and Zadro, 2005). As such, applying behavioural strategies is often ineffective in mitigating the undesirable effects of ostracism on individuals; indeed, in some cases, inappropriate strategies may even make the situation worse. Researchers have long sought to understand how and why influence tactics are effective, and their lack of understanding has created a barrier to realizing the nature of the processes of social influence (Jones, 1990). This is particularly the case for ingratiation, which requires the actor to have an adequate level of ability to perform properly in the workplace (Treadway et al., 2007).
Political skill is an important ability for the actor when exhibiting ingratiating behaviours (Treadway et al., 2007). Early research defined political skill as the ability to exercise influence through the use of influence tactics (Mintzberg, 1983). More recently, political skill has been defined as an individual's ‘ability to effectively understand others at work and to use such knowledge to influence others to act in ways that enhance one's personal and/or organizational objectives’ (Ferris et al., 2005, p. 127). People with high levels of political skill are more likely to pick up on social cues and understand the motivations of others. Scholars have suggested that political skill may overlap with, but is sufficiently different from, other related constructs (Ferris et al., 2007). For example, the distinction between ingratiation and political skill is that ingratiation is social influence behaviour, whereas political skill is an ability that employees have when displaying that behaviour. A recent study by Treadway et al. (2007) demonstrated that political skill relates to ingratiation at a modest level (r = 0.14, p < 0.05). Another social competency construct regarded as being important in identifying political skill is that of emotional intelligence, which involves facilitating interpersonal behaviour (Ferris et al., 2007). The primary difference is that emotional intelligence sheds light on the emotion-based aspects of interpersonal effectiveness, influence, and control, whereas political skill is conceptualized as incorporating skills and knowledge that go beyond the emotions (for a review, see Ferris et al., 2005, 2007).
Accordingly, the purpose of this study is to examine the relationship between workplace ostracism and employee psychological distress, by focusing on the joint moderating effects of ingratiation and political skill. We propose that political skill is an important variable with respect to the success and failure of using ingratiation to cope with workplace ostracism. In other words, an actor's political skill may explain why that actor, successfully or unsuccessfully, applies ingratiation to counter ostracism, thereby making the situation better or worse. More specifically, when an employee's political skill is high, ingratiation neutralizes the relationship between workplace ostracism and psychological distress; when employee political skill is low, ingratiation exacerbates the relationship.
This study makes three major contributions to the literature related to workplace ostracism, psychological distress, ingratiation, and political skill. First, we extend both the workplace ostracism and psychological distress literature by theoretically and empirically testing the ostracism model (Williams, 1997, 2001) to connect workplace ostracism and psychological distress in a field setting. Second, we examine the joint moderating roles of a coping behaviour (i.e. ingratiation) and an individual ability (i.e. political skill), providing boundary conditions for the relationship between workplace ostracism and psychological distress.
Finally, we use a two-wave design to test the proposed hypotheses among employees in the People's Republic of China (workplace ostracism, ingratiation, and political skill were measured at time 1, while psychological distress was measured at time 2). This further contributes to the ingratiation and political skill literature, in which cross-sectional design predominates (Treadway et al., 2007). Using Chinese samples also helps to test Western theories and findings in the Chinese context, and responds to calls for non-Western samples in ingratiation and psychological distress research (Aryee et al., 1996; Cole et al., 2010). The Chinese tend to have various strong interpersonal ties and high levels of collectivistic values; as such, establishing and maintaining close relationships with others at work represents personal value and career success to Chinese employees (Liu et al., 2009). Research suggests that individuals who hold interdependent views of the self are more likely to experience negative emotions when negative events happen (Lee and Tiedens, 2001). Hence, Chinese cultural characteristics may augment the negative influence of workplace ostracism on the sense of self of an individual, making China an ideal setting for researchers to examine this influence. In practical terms, China has strong business ties with Western countries, as evidenced by the fact that China receives the highest amount of foreign direct investment (Peng, 2006). In addition, China is home to a high percentage of the world's population, with more than one billion inhabitants, and Chinese employees' participation in the global economy has grown sharply (Xie et al., 2008). All of these aspects emphasize the importance of conducting management studies using Chinese respondents.
THEORETICAL BACKGROUND AND HYPOTHESES
Workplace Ostracism and Employee Psychological Distress
As an interpersonal stressor, ostracism threatens the social resources of the target, which are assets that can be drawn upon when needed, to solve a problem or cope with a challenging event (Greenhaus and Powell, 2006). Based on the conservation of resources (COR) theory, people strive to retain, protect, and establish resources, given that such resources are limited (Hobfoll, 1989, 2001). They thus find it threatening when they see a potential or substantial loss of these valuable resources. Indeed, resource loss events are responsible for most cases of depression (Hobfoll, 1989). Ostracism presents significant challenges that can decrease the resources that individuals can hold. This is because, on the one hand, individuals need to mobilize resources to counter ostracism, and on the other hand, they are less likely to refill their resources from other people, leading to a situation in which resources are drained away. As resources can support individuals in handling their daily work, people who run out of resources are likely to become stressed and exhausted (Hobfoll, 1989).
Williams (1997, 2001) developed the most popular model to predict ostracism outcomes. This model is premised on the belief that ostracism can threaten social resources, and consequently can be viewed as a stressor (Williams, 2001). Recent studies have demonstrated that social ostracism leads to a series of psychologically aversive reactions, including anger (Chow et al., 2008) and negative mood (Gonsalkorale and Williams, 2007). In particular, organizational research has revealed that workplace ostracism is related to higher levels of anxiety, depression, job search behaviour, and turnover intentions, as well as lower levels of satisfaction and psychological health (Ferris et al., 2008; Hitlan et al., 2006). Interestingly, experimental studies provide evidence that the source of ostracism does not significantly moderate the negative effects of ostracism on the target of being ostracized. One study demonstrated that people who are ostracized by in-group and out-group members manipulated by two groups of computer users (PC and Macintosh users) suffer similar levels of loss of belonging (Williams et al., 2000). A further study revealed that being ostracized by a despised out-group member is no less detrimental than being ostracized by rival out-group and in-group members (Gonsalkorale and Williams, 2007).
Although all of the previously described consequences are undesirable, this study focuses on employees' psychological distress. Such distress includes job tension, emotional exhaustion, and depressed mood at work, all of which have been linked to important job, family, and health outcomes such as job satisfaction, work–family conflict (Grandey et al., 2005), organizational citizenship behaviour, workplace deviance (Lee and Allen, 2002), job performance (Cropanzano et al., 2003), intention to leave (Harvey et al., 2007), and high blood pressure (Schaubroeck and Merritt, 1997).
Job tension has long been defined as ‘the psychological reaction of workers to disturbances in the objective or perceived work environment’ (Chisholm et al., 1983, p. 387). Scholars have also applied this definition in recent research (Harvey et al., 2007). Job tension, referring to the effect of stressful conditions in the workplace (Pool, 2000), is an important mediator between job stressor and job performance (Hochwarter et al., 2007) and is included in many studies to reflect the stress effect (Klenke-Hamel and Mathieu, 1990). Job tension is distinguishable from job stress in that the latter stems from environmental situations or events potentially capable of providing a state of stress (e.g. work overload) (Greenhaus and Parasuraman, 1986), whereas the former is a psychological reaction to job stress, and occurs within the body in response to job stress (Macan, 1994). As the present study focuses on psychological distress, job tension rather than job stress is identified as an outcome.
In extending Williams' (1997) model that links ostracism and threatened needs, we argue that ostracism leads to the depletion of employees' capacity to maintain resources to meet demand and fulfil expectations, which in turn tends to create job tension between one's self and the workplace environment (i.e. job demands and performance expectations). This suggests a positive relationship between workplace ostracism and job tension.
Moreover, the most widely reported component of burnout is emotional exhaustion, defined as ‘a chronic state of emotional and physical depletion’ (Cropanzano et al., 2003, p. 160). This involves feelings of being overextended and depleted of one's emotional and physical resources (Halbesleben and Bowler, 2005). Emotional exhaustion occurs when emotional demands exceed what individuals can afford during interpersonal interactions (Maslach et al., 2001).
Ostracism constitutes a form of resource loss in terms of work support from colleagues. According to research, when individuals feel that they do not have enough resources to handle the daily work that confronts them, they experience emotional exhaustion (Lee and Ashforth, 1996). When employees are ostracized, their emotional connections with others are cut. Humans need social interaction as a means of sharing their emotional feelings, in order to enhance their emotional resources and maintain their psychological and physical health (Heaphy and Dutton, 2008). When the needs of emotional sharing cannot be fulfilled, emotional resources are lost, thereby leading to emotional exhaustion. This suggests a noteworthy relationship between workplace ostracism and emotional exhaustion.
Another form of psychological strain is that of depressed mood, which refers to feeling blue and downhearted (Prati et al., 2009). Ostracism is related to depressed mood for two reasons. First, ostracism causes a painful and negative experience (Gruter and Masters, 1986). Research has revealed that negative experiences cause intense emotional reactions, including strain and distress (Taylor, 1991). Second, facing a stressful work environment alone can increase distress. Studies using experiments and surveys have provided evidence that the target of ostracism displays a depressed mood (Ferris et al., 2008; Leary, 1990; Williams et al., 2002).
Overall, workplace ostracism is regarded as a workplace stressor that threatens employees' resources that enable them to cope with their work and daily life. Following the theory of COR (Hobfoll, 1989, 2001) and the model of ostracism (Williams, 1997, 2001), we propose:
Hypothesis 1: Workplace ostracism is positively related to (a) job tension, (b) emotional exhaustion, and (c) depressed mood at work.
Joint Moderating Effects of Ingratiation and Political Skill
Ingratiation is commonly used to handle ostracism (Williams and Zadro, 2005). Research has indicated that when effectively implemented, ingratiation can provide desired images for individuals (Judge and Bretz, 1994) and can encourage a friendly work environment that promotes favourable social interactions (Harvey et al., 2007). Social resources that are facilitated by social interactions could serve to buffer the adverse effects of stressors on outcomes. Research has delineated two ways in which the buffering mechanisms of social resources occur (Cohen and Wills, 1985). First, social resources may influence the relationship between the stressful event and the corresponding stress reaction by depressing or preventing a stress appraisal response. Second, social resources may weaken the association between the stress experience and the outcome by suggesting a solution to the problem. Hence, social resources can attenuate the relationships between job stressors and negative outcomes, implying a moderating effect.
However, displaying ingratiation may not fully attenuate the negative consequences related to workplace ostracism for all individuals. Considerable research has suggested that coping responses differ among individuals (Williams, 2007). Indeed, the success of ingratiation depends on whether it is effectively implemented. Research has shown that ingratiatory behaviours perceived as being driven by impression management motives may lead to negative outcomes, such as low performance ratings and poor promotability (Bolino, 1999). Studies have also revealed that when unsuccessfully implemented, ingratiation can cause unintended negative images, such that the influencer is regarded as a sycophant when the influencee detects the ingratiation and suspects its underlying motives (Lam et al., 2007; Turnley and Bolino, 2001). Thus, for ingratiation to be effective, the employee needs to exhibit ingratiation appropriately, which requires skill.
To explain the divergent effects of ingratiation, recent studies have focused on the political skill of the influencer when investigating the ingratiation–outcomes linkage (Harris et al., 2007; Treadway et al., 2007), based on the rationale that the effectiveness of ingratiation is ‘a function of employee political skill, which facilitates the delivery and execution of the influence behaviour’ (Ferris et al., 2007, p. 309). Ferris et al. (2005) contend that political skill includes four dimensions associated with ingratiation effectiveness: social astuteness, interpersonal influence, networking ability, and apparent sincerity. Employees who are socially astute and master interpersonal influence are more adept at changing their attitudes, perspectives, and behaviours to fit in with the environment (Ferris et al., 2005; Turnley and Bolino, 2001). When applying ingratiation, these employees know how to fulfil their goals successfully. Moreover, politically skilled employees understand how to develop and apply social networks (Ferris et al., 2005). This networking ability helps them to establish high-quality relationships with their colleagues. In such a context, ingratiation attempts are likely to be more effective (Ferris et al., 2007). Finally, politically skilled employees can influence and manage others' responses when appearing to have high levels of integrity, sincerity, and genuineness (Ferris et al., 2005). Overall, ingratiation attempts displayed by employees with high levels of political skill are more likely to be effective.
Consistent with these arguments, empirical findings have demonstrated that the influencer's political skill is a key determining factor in the success and failure of ingratiation attempts. For instance, in using samples in a state government agency, Harris et al. (2007) found that ingratiation was positively related to supervisors' ratings of job performance among employees high in political skill, but was negatively related among those low in political skill. In addition, results from data drawn from two retail service organizations indicated that ingratiation was less likely to be detected by the supervisor when employees were politically skilful (Treadway et al., 2007).
Taken together, the influencer's political skill clearly impacts the ingratiation effectiveness in relation to coping with ostracism, such that employees high in political skill are more capable of displaying ingratiation effectively than those low in political skill. According to existing research, employees who impact others' opinions of themselves successfully can separate themselves from job tension and emotional exhaustion, thereby experiencing less psychological distress (Harvey et al., 2007). Conversely, employees who display high levels of ingratiation with low levels of political skill may experience more psychological distress than employees who display low levels of ingratiation when faced with workplace ostracism, as ingratiation is more likely to be regarded as self-serving and manipulative when the performer holds a low degree of political skill (Treadway et al., 2007). Empirical findings provide evidence that failure to use coping measures to counter the stressor leads employees to experience more job strain (Zellars et al., 2004). Thus, we propose:
Hypothesis 2: The moderating effect of ingratiation on the relationships between workplace ostracism and (a) job tension, (b) emotional exhaustion, and (c) depressed mood at work depends on the level of employee political skill. At higher levels of political skill, workplace ostracism has a less positive relationship with job tension, emotional exhaustion, and depressed mood at work among employees reporting high rather than low ingratiation. At lower levels of political skill, workplace ostracism has a more positive relationship with job tension, emotional exhaustion, and depressed mood at work among employees reporting high rather than low ingratiation.
Sample and Procedures
The participants in this study were the employees of two large oil and gas firms located in a northwest city of China. Two waves of data collection were conducted in order to reduce the potential common method bias (Podsakoff et al., 2003). In the first-wave survey (T1), the employees provided information on political skill, ingratiation behaviour, perceptions of workplace ostracism, and control variables. In the second-wave survey (T2), conducted four months after T1, participants were asked to report on their perceptions of job tension, emotional exhaustion, and depressed mood at work. The scales were translated into Chinese, following the commonly used back translation procedure (Brislin, 1980).
Human resource managers from the firms prepared a list of 320 randomly selected employees. Survey questionnaires were coded before distribution, and the human resources departments assisted in matching the identity numbers and names of respondents with the responses of the T1 and T2 surveys conducted. Questionnaires were then administered to the selected employees. The employees acquired the survey packets in a meeting room during work hours, in the presence of data collectors trained and led by one of the authors. Respondents were informed that the survey aimed to examine the experience of human resource practices; they were assured of the confidentiality of their responses. Each respondent placed his/her completed survey into a sealed envelope, and returned it to a box positioned in the human resources department.
In T1, 268 questionnaires were completed and returned, representing a response rate of 83.8 per cent. Four months later, when the T2 survey was conducted, questionnaires were distributed to these 268 employees, with 215 questionnaires completed and returned, representing a response rate of 80.2 per cent. Hence, the final sample of this study consisted of 215 employees. Of these employees, 51.2 per cent were male and 71.2 per cent were skilled technicians. The remaining 28.8 per cent were administrative staff. The average age was 36.36 years (SD = 6.36) and the average number of years of tenure with the organization was 9.97 years (SD = 5.08). The jobs of these employees involved relatively high levels of social interactions. Thus, it appears that this sample is relevant and suitable for testing our theoretical model.
Workplace ostracism. A ten-item scale developed by Ferris et al. (2008) was used to measure workplace ostracism. Response options ranged from 1 (‘strongly disagree’) to 5 (‘strongly agree’). Sample items included ‘Others ignored me at work’, ‘Others left the area when I entered’, and ‘My greetings have gone unanswered at work’. Cronbach's alpha for this measure was 0.90.
Ingratiation. Ingratiation was assessed using a modified four-item subscale taken from Bolino and Turnley's (1999) impression management scale. Response options ranged from 1 (‘never behave this way’) to 5 (‘often behave this way’). Items included ‘Praise others at work for their accomplishments so they will consider me a nice person’, ‘Do personal favours for others at work to show them that I am friendly’, ‘Compliment others at work so they will see me as likeable’, and ‘Take an interest in others' personal lives to show them that I am friendly’. Cronbach's alpha for this measure was 0.87.
Political skill. Ferris et al.'s (2005) 18-item Political Skill Inventory was used to assess political skill. Response options ranged from 1 (‘strongly disagree’) to 5 (‘strongly agree’). Sample items included ‘I pay close attention to people's facial expressions’, ‘It is easy for me to develop good rapport with most people’, and ‘I am able to communicate easily and effectively with others’. Confirmatory factor analysis (CFA) of the 18 items produced good fit indices (χ2 (129) = 234.576, p < 0.01; CFI = 0.95, TLI = 0.94; RMSEA = 0.062). Cronbach's alpha for the whole measure was 0.93.
Job tension. A seven-item scale developed by House and Rizzo (1972) was used to measure job tension. Response options ranged from 1 (strongly disagree) to 5 (strongly agree). Sample items included ‘I work under a great deal of tension’, ‘I have felt fidgety or nervous as a result of my job’, and ‘My job tends to directly affect my health’. Cronbach's alpha for this measure was 0.88.
Emotional exhaustion. A five-item scale developed by Schaufeli et al. (1996) was used to measure emotional exhaustion. Response options ranged from 1 (‘strongly disagree’) to 5 (‘strongly agree’). Sample items included ‘I feel emotionally drained from my work’ and ‘I feel burned out from my work’. Cronbach's alpha for this measure was 0.91.
Depressed mood at work. A ten-item scale developed by Quinn and Shepard (1974) was used to measure employees' depressed mood at work. Response options ranged from 1 (‘strongly disagree’) to 5 (‘strongly agree’). Sample items included ‘I feel downhearted and blue’, ‘I get tired for no reason’, and ‘I find myself restless and can't keep still’. Cronbach's alpha for this measure was 0.94.
Control variables. Following previous research (e.g. Harvey et al., 2007; Hochwarter et al., 2005; Wu and Hu, 2009; Zellars et al., 2004), we controlled for employee age, gender, organizational tenure, occupation, and negative affectivity, all of which are likely to be associated with job tension, emotional exhaustion, and depressed mood at work. Age and organizational tenure were self-reported in years. Gender was dummy-coded, with male coded as ‘0’ and female coded as ‘1’. Occupation was also dummy-coded, with administrative staff coded as ‘0’ and technicians coded as ‘1’. Negative affectivity was measured using a ten-item scale developed by Watson et al. (1988). Sample items included ‘irritable’, ‘nervous’, and ‘upset’. Cronbach's alpha for this measure was 0.92.
Confirmatory Factor Analysis
Confirmatory factor analyses (CFAs) were conducted to ensure sufficient convergent and discriminant validity among all constructs. Given the small sample size relative to the measurement items, we adopted procedures frequently used by researchers (Hui et al., 2004). First, we reduced the number of items by creating three indicators for each single-dimension construct. Based on the factor analysis results, the items with the highest and lowest loadings for each construct were combined first, followed by the items with the next highest and lowest loadings, until all the items had been assigned to one of the indicators. Scores for each indicator were then computed as the mean of the scores on the items that constituted each indicator. Second, for political skill, which is a four-dimension construct, we reduced the number of items by creating four indicators, with each indicator being represented by the dimension score.
We first examined a seven-factor CFA model that included negative affectivity, workplace ostracism, political skill, ingratiation, job tension, emotional exhaustion, and depressed mood at work. The overall model's chi-squared, comparative fit index (CFI) (Bentler, 1990), root mean square error of approximation (RMSEA) (Browne and Cudeck, 1993), and Tucker–Lewis Index (TLI) (Tucker and Lewis, 1973) were used to assess model fit. The proposed seven-factor model fitted the data well (χ2 (188) = 211.65, p < 0.01; CFI = 0.99, TLI = 0.99; RMSEA = 0.024). In addition, all factor loadings were significant, demonstrating convergent validity.
The discriminant validity of the seven proposed constructs was tested by contrasting the seven-factor model against alternative models. Model comparison results (see Table I) revealed that the seven-factor model fitted the data considerably better than any of the alternative models. Thus, the distinctiveness of the seven constructs in the study was supported. Given these results, all seven constructs were applied in subsequent analyses.
Table I. Results of confirmatory factor analysis for the measures of the variables studied
Notes: TLI is the Tucker–Lewis index; CFI the comparative fit index; and RMSEA the root-mean-square error of approximation.
p < 0.05;
p < 0.01.
Six-factor model 1: Workplace ostracism and job tension combined
As the data were collected from the same source, common method variance was a likely concern. Several procedural and statistical remedies suggested by Podsakoff et al. (2003) were used to minimize potential common source bias. First, the participants were assured of the anonymity and confidentiality of responses in order to reduce issues such as participants' evaluation apprehension and social desirability. Second, a psychological separation in the survey was constructed using different instructions and interspersing all variables throughout the survey, with a number of filler items included among them. This step aimed to lower respondents' perception of any direct connection between these variables. Third, a four-month interval was included between the data collection of the predictors and the consequences. Fourth, we statistically tested the potential influence of common method variance using Harman's one-factor test. Principal factor analysis with varimax rotation was performed to determine whether a single method factor explained a majority of variance. Several factors with eigenvalues greater than one were reported, with the first factor accounting for 18.98 per cent of the total variance explained. Finally, we calculated the variance explained by the method factor (12.11 per cent); the result was much lower than the 25 per cent suggested by Williams et al. (1989). Hence, common method variance is unlikely to be a pervasive problem in this study.
Table II presents the means, standard deviations, and zero-order Pearson correlations of all key variables. As indicated, workplace ostracism was found to be positively correlated with job tension (r = 0.25, p < 0.01), emotional exhaustion (r = 0.29, p < 0.01), and depressed mood at work (r = 0.26, p < 0.01).
Table II. Means, standard deviations, and correlations
Notes: Values in parentheses on the diagonal are the Cronbach's alpha value of each scale.
We followed Cohen et al.'s (2003) procedures by conducting a five-step hierarchical multiple regression analysis to test our hypotheses. The control variables (employee age, gender, organizational tenure, occupation, and negative affectivity) were entered first, followed by workplace ostracism in the second step. Ingratiation and political skill were entered in the third step. Three two-way interaction terms (workplace ostracism × political skill; workplace ostracism × ingratiation; political skill × ingratiation) were entered in the fourth step. Finally, a three-way interaction term (workplace ostracism × political skill × ingratiation) was entered in the fifth step for predicting employee psychological distress (job tension, emotional exhaustion, and depressed mood at work). The variables used in the interaction terms were centred to reduce any multicollinearity (Aiken and West, 1991).
Table III presents the results of the hierarchical multiple regression analysis. Hypothesis 1 predicted that the relationships between workplace ostracism and job tension, emotional exhaustion, and depressed mood at work were positive. The results demonstrated that workplace ostracism was positively related to job tension (β = 0.22, p < 0.01), emotional exhaustion (β = 0.26, p < 0.01), and depressed mood at work (β = 0.24, p < 0.01), supporting Hypotheses 1a, 1b, and 1c.
Table III. Results of hypotheses testing
Job tension (T2)
Emotional exhaustion (T2)
Depressed mood at work (T2)
Notes: WOS = workplace ostracism; PS = political skill; IN = ingratiation.
Hypothesis 2 proposed that ingratiation and political skill jointly moderate the relationship between workplace ostracism and psychological distress (job tension, emotional exhaustion, and depressed mood at work). As shown in Table III, the three-way interaction term proved to be significantly related to job tension (β = −0.19, p < 0.01), emotional exhaustion (β = −0.17, p < 0.05), and depressed mood at work (β = −0.17, p < 0.05), supporting Hypotheses 2a, 2b, and 2c.
The nature of the significant three-way interactions was examined by plotting values plus and minus one standard deviation from the means of workplace ostracism, ingratiation, and political skill (Cohen et al., 2003). Figures 1–3 clearly illustrate the significant three-way interactions. The upper panels of these three figures demonstrate the interactive effect of workplace ostracism and ingratiation on psychological distress (job tension, emotional exhaustion, and depressed mood at work) for employees high in political skill. When employees performed high levels of ingratiation, workplace ostracism was unrelated to job tension (β = −0.11, n.s.), emotional exhaustion (β = −0.03, n.s.), and depressed mood at work (β = −0.09, n.s.). In contrast, when employees performed low levels of ingratiation, workplace ostracism was positively related to job tension (β = 0.26, p < 0.01), emotional exhaustion (β = 0.31, p < 0.01), and depressed mood at work (β = 0.25, p < 0.01).
Meanwhile, the lower panels of Figures 1–3 demonstrate the interactive effect of workplace ostracism and ingratiation on psychological distress for employees low in political skill. When employees performed high levels of ingratiation, workplace ostracism was more positively related to job tension (β = 0.56, p < 0.01), emotional exhaustion (β = 0.55, p < 0.01), and depressed mood at work (β = 0.57, p < 0.01). In contrast, when employees performed low levels of ingratiation, workplace ostracism was less positively related to job tension (β = 0.18, p < 0.05), emotional exhaustion (β = 0.21, p < 0.01), and depressed mood at work (β = 0.23, p < 0.01). These patterns provide additional support for Hypothesis 2.
The survey results support all the hypotheses by demonstrating that the relationship between workplace ostracism and psychological distress is moderated by ingratiation and political skill. However, a major limitation is associated with the survey study: all of the ingratiation items relate to the general application of ingratiation, such that it is impossible to determine whether people respond to ostracism using ingratiation. Hence, using a critical incident approach, we conducted interviews to ask respondents to provide a qualitative description of their most recent experience of ostracism and how they handled it. Given time and resource constraints, we randomly selected and approached 22 employees who had completed both T1 and T2 survey questionnaires. As the original 215 respondents were coded as numbers for the survey study, we put their numbers in a box and randomly drew 22 numbers from the box. Based on the code sheets that recorded the identity numbers and respondents' names, we identified those 22 employees and asked them whether they were willing to be interviewed. Of those 22 randomly selected employees, 15 agreed to be interviewed and described their methods to counter workplace ostracism. Seven of the 22 employees selected were not willing to be interviewed, indicating that they were busy at work and could not afford the time. Each interview lasted from 10 to 30 minutes. In the final few interviews, we were unable to find any new information and therefore decided not to add any other respondents, resulting in 15 interviewees. To examine the representativeness of the interviewees, we compared their ingratiation, psychological distress, and political skill levels to those of the entire sample responses and did not find any significant differences. Examples of what they had experienced included: ‘My colleagues refused to talk to me and answer my questions’ and ‘My colleagues did not ask me whether I would like to join them for lunch’. Two doctoral students who were unaware of the goals of the present study independently coded the descriptions based on nine themes developed from the study of Yukl et al. (1993). The themes include ingratiation, rational persuasion, inspirational appeals, consultation, personal appeals, exchange, coalition tactics, pressure, and legitimating tactics. Agreement between the two coders was over 95 per cent. In situations where disagreement occurred, a consensus was reached through discussion. The results indicate that ingratiation, rational persuasion, and personal appeal are the most frequently used methods, and that 14 out of 15 interviewees mentioned the application of ingratiation to counteract workplace ostracism, which is consistent with past findings that ingratiation is a common method for counteracting ostracism (Williams and Zadro, 2005). These results serve to alleviate concerns as to the general nature of the measurement of ingratiation.
The findings of this study revealed support for the hypotheses regarding the relationships between workplace ostracism, ingratiation, political skill, and psychological stress (job tension, emotional exhaustion, and depressed mood at work). First, workplace ostracism was found to be positively related to psychological distress. Second, we found that the relationship between workplace ostracism and psychological distress depended on the levels of employee ingratiation and political skill. High ingratiation coupled with high political skill resulted in the weakest relationships of workplace ostracism with job tension, emotional exhaustion, and depressed mood at work. Conversely, high ingratiation, combined with low political skill, led to the strongest positive relationships of workplace ostracism with the outcomes.
Although similar to prior findings regarding the positive relationship between workplace ostracism and psychological distress consequences (Ferris et al., 2008), our findings indicated that these relationships were moderated by ingratiation and political skill. This suggests that solely exhibiting ingratiation cannot in itself attenuate the effects of workplace ostracism on psychological distress, and that political skill should be considered when the employee engages in ingratiation. In other words, the employee should also possess high levels of political skill when deploying ingratiation to counteract ostracism. When the employee possesses a low level of political skill, using ingratiation may make the situation worse, and increase psychological distress.
This study makes several contributions to the growing literature on workplace ostracism, ingratiation, and political skill. First, we expanded the nomological net of workplace ostracism by examining the moderators of behavioural strategy (i.e. ingratiation) and individual ability (i.e. political skill) in a single study. Our findings, combined with our literature review, have provided evidence that when the target engages in ingratiation to cope with workplace ostracism, high political skill is needed to make other people categorize the behaviour as prosocial, rather than self-serving and manipulative, engendering a positive impact on the target. This extends Williams' (1997, 2001) model by providing moderators for the relationship between ostracism and psychological distress. Although Williams' (1997, 2001) model focuses on individual short-term and long-term responses to ostracism, no explanation is provided as to why some responses work while others do not, because the model does not take moderators into consideration. We can conclude that the target's ability, combined with behavioural strategies, should be included in the model.
By extending this theoretical implication, it is also possible that individual factors such as ingratiation and political skill may be used to cope not only with workplace ostracism, but also with other destructive behaviours such as abusive supervision. Research has shown that employees who face abusive supervision have higher levels of psychological distress (Tepper, 2000). Although not specifically investigated in this study, it is plausible that effective ingratiation coupled with high levels of political skill can help employees cope with destructive behaviours and maintain psychological health. Hence, ingratiation and political skill can be viewed as important moderators in extending the existing coping strategy literature.
In practical terms, psychological distress is costly for both managers and organizations (Harvey et al., 2007). Our work suggests two paths by which managers and organizations can improve levels of employee psychological distress. The first is to take steps to reduce workplace ostracism. Williams' (1997, 2001) model proposes that individual differences and social/situational forces may facilitate the use of ostracism. As for individual differences, employees who have experienced ostracism in the past, and hold low levels of self-esteem, tend to apply ostracism. Therefore, managers need to decrease their use of ostracism, and organizations need to create a culture that discourages workplace ostracism when providing training for both managers and employees to enhance self-esteem, and avoid the use of ostracism. In training sessions, organizations can give advice to their employees about how the effects of ostracism are far more harmful than imagined, and suggest that employees solve any problems through discussion (Williams, 2001).
The second path to reducing employee psychological stress is to promote ingratiation tactics, which are an effective behavioural strategy for employees who have high levels of political skill. Managers and organizations that want to reduce employee psychological distress should recruit employees with high levels of political skill and promote political skill through training, counselling, and mentoring. Moreover, managers and organizations can encourage employees with low levels of political skill to use other behavioural strategies to cope with workplace ostracism. Appropriate strategies may include mediation, clarity seeking, and forgiveness seeking (Williams and Zadro, 2005).
Several limitations of the present research should also be noted. First, as we were constrained by the length of the questionnaires, other important behavioural strategies (e.g. rational persuasion and personal appeal) and potential control variables such as personality hardiness and need for affiliation were not included in this study. We are not certain about the effects of workplace ostracism, ingratiation, and political skill when these behavioural strategies and personality variables are included. Indeed, individuals may simultaneously use more than one strategy to cope with ostracism (Williams, 2001), and personality variables lead people to be more or less sensitive to stressors (Judge et al., 2009). Therefore, it may be necessary to incorporate additional variables related to behavioural strategies and personality variables to investigate how far workplace ostracism, ingratiation, and political skill can account for psychological distress outcomes.
Second, because of the four-month interval research design, our study focused on long-term outcomes while ignoring short-term outcomes, such as the emotions. Previous research has indicated that ostracism can cause anger and sadness, leading to predicted behaviours (Chow et al., 2008) and decision making (Lerner and Tiedens, 2006). Hence, further studies should use daily and within-subject research designs, to examine how workplace ostracism influences employees over a short time period.
Third, although the measurement of workplace ostracism applied in this study is psychometrically sound, it does not distinguish between the types and sources of workplace ostracism. Some people may experience more of one kind of ostracism than another (e.g. intentional versus unintentional), which consequently influences their use of ingratiation and political skill as well as the effectiveness of their coping mechanism. Moreover, it is possible that power difference affects workplace ostracism and ingratiation to the extent that workplace ostracism is more severe if it comes from supervisors than from peers and subordinates, as people in low power roles are more sensitive to other individuals' thoughts than high power ones (Lee and Tiedens, 2001). With regard to the effect of ingratiation, research has long suggested that ingratiation exhibited in high power roles has a greater impact than that exhibited in low power roles (Aryee et al., 1996). Moreover, ingratiation targeted at supervisors may help counteract the negative effects of ostracism from peers and subordinates, providing a cross-domain buffering effect. Thus, future research should develop a multi-dimensional scale to capture different types and sources of workplace ostracism, and examine the effects of power differences and cross-domain buffering.
Fourth, our findings may be contaminated by common method variance related to data on workplace ostracism, job tension, emotional exhaustion, and depressed mood at work from the same source. However, as previously discussed, based on the suggestions of Podsakoff et al. (2003), we applied several methods to attenuate common method variance. Moreover, the presence of common method bias does not necessarily influence results (Spector, 2006). Therefore, we believe that common method variance does not impose a great threat to our findings. Nevertheless, future research could use other ratings (e.g. supervisors and peers) to alleviate this potential bias.
Finally, a constraint on the generalizability of our findings stems from the use of Chinese data. Although China provides an ideal setting in which to examine the stated hypotheses, our findings may not be generalizable to Western nations. Past research has revealed that interdependence (interdependent self-construal) strengthens the relationship between ostracism and negative consequences (Lee and Tiedens, 2001), which may imply that this relationship could be weaker in the West. Furthermore, prior studies have suggested that power distance renders ingratiation less effective in China than in the United States (Aryee et al., 1996), which may suggest that ingratiation in the West could provide stronger effects than in China. Hence, it is desirable to replicate our research by considering cultural characteristics and using Western samples.
Despite these limitations, this study has addressed crucial issues regarding workplace ostracism, ingratiation, political skill, and psychological distress. It has revealed that ingratiation and political skill are key moderators that can further our understanding of the relationships of workplace ostracism linked with job tension, emotional exhaustion, and depressed mood at work. We hope that the study offers a springboard for future research on workplace ostracism, ingratiation, political skill, and psychological distress.
This work was supported by a grant from China's National Natural Science Foundation (70802060) and a joint grant from China's National Natural Science Foundation and Hong Kong Research Grants Council (71061160504-HKU778/10). A previous version of this article received the Best Micro Paper Award at the 2010 International Association for Chinese Management Research Conference, Shanghai, China. The authors would like to thank Jia Lin Xie for her helpful comments on early drafts of this paper.