Although the topic of workplace bullying was in its infancy a decade ago (Sheehan, 1999), research regarding the construct has recently grown (McMahon, 2000; Agervold and Mikkelsen, 2004; Vickers, 2009; Hutchinson and Eveline, 2010). Amid worldwide competition, uncertainty and pressure inherently pervade contemporary organisations (Sheehan, 1999). In such situations, people in many organisations realise that they are being bullied (Sheehan and Jordan, 2000). Workplace bullying is an important issue, as it occurs globally in organisations (Saam, 2010; Mathisen et al, 2011). At the same time, it is a sad reality (Pate and Beaumont, 2010) that has negative outcomes similar to workplace abuse, violence or harassment (Sheehan and Jordan, 2000). Hence, it has been termed the ‘silent epidemic’ (Namie and Namie, 2009).

The detrimental consequences of bullying identified in the literature are decreased work satisfaction among employees (Uppal, 2005), increased employee turnover and lack of morale in employees. Such consequences have unfavourable outcomes for organisational performance. Workplace bullying is particularly deleterious for staff, patients and outcomes (Michelle et al, 2010) in health-care organisations. Hence, it is important to investigate factors that may lessen the expected adverse impact of bullying behaviour at work, with the aim of enhancing job performance and subsequently organisational performance.

Job performance is an organisational variable that managers and HR practitioners aim to maximise. Concurrently, EI has been recognised as an ability that ensures success at the organisational level. Using one’s emotional ability to cope with bullying perpetrated by managers is likely to reduce the negative effects that workplace bullying might have on employee job performance. Previous literature has identified that managers bully subordinates because they are unable to react emotionally in an appropriate way (Sheehan and Jordan, 2000). Other studies have focused on bullying behaviour as a consequence of low EI in the perpetrator (Mathisen et al, 2011). The present study is unique in that it focuses on the EI of those bullied at work. Moreover, the scheme presented in this study corresponds with Ashforth and Humphrey’s (1995) call for organisational research to include emotional variables. In sum, whether the EI of those bullied at work impacts their job performance is currently unknown; this is an important gap the present study aims to bridge. Regarding the health-care industry, the construct of workplace bullying has been tested among nurses (Hutchinson and Eveline, 2010; Lindy and Schaefer, 2010), yet remains barely investigated among doctors, despite the proposition that bullying is rife in this occupation.

Clearly, the delineated relationships require more precise and rigorous attention. This study empirically examines the possible moderating role of EI in the relationship between workplace bullying and job performance. Hence, it proposes a holistic framework and theory advancement.

Literature Review

Workplace bullying

Certain scholars (McMahon, 2000; Sheehan and Jordan, 2000; Lindy and Schaefer, 2010; Pate and Beaumont, 2010; Stogstad et al, 2011) have categorised workplace bullying using various terms, for example, mobbing, emotional abuse, occupational stress, workplace violence, workplace abuse and workplace harassment. In contrast, others (Vickers, 2009; Pontzer, 2010; Mathisen et al, 2011) have not differentiated between bullying and similar terms. Nonetheless, these terms share commonality. Workplace bullying is essentially psychological violence (Namie and Namie, 2009) that is manifest in targeted acts of belittlement, such as intimidation and humiliation, criticism of abilities and competence (Brotheridge and Lee, 2010) and invention of mistakes (Namie, 2003). Workplace bullying also includes acts that undermine targets’ work, such as refusing relevant information, coercing work and setting difficult goals (Cheema et al, 2005; Brotheridge and Lee, 2010). Such behaviour can result in drinking problems, depression, stress and career dissatisfaction (Frank et al, 2006), as well as psychological distress (Tepper, 2007) in targets. Throughout this study, we refer to the study variable as workplace bullying and operationalise it as inclusive of the stated negative acts. ‘Bullying at work means harassing, offending, socially excluding someone or negatively affecting someone’s work tasks’ (Einarsen et al, 2003, p. 15).

Managers may resort to bullying as a means to achieve results, as they are pressurised to attain organisational goals (Sheehan and Jordan, 2000) using whatever means (Sheehan, 1999). Following a manager’s hostility, or even inappropriate jokes meant to relieve stress, employees may perceive that they are being bullied (Sheehan and Jordan, 2000). Bullying is considered as repetitive and regular behaviour (Saam, 2010; Mathisen et al, 2011) aimed at one or more organisational members, and excludes rare flare-ups (Stogstad et al, 2011).

Workplace Bullying and Job Performance

The consequences of bullying on those targeted are invariably destructive (Vickers, 2009). Past research has associated workplace bullying with serious detrimental outcomes for targets and indicates that it jeopardises individual (Namie and Namie, 2009) and organisational performance (McMahon, 2000; Fernández-Berrocal and Extremera, 2006; Townend, 2008). Bullying also results in absences and inclinations to quit (Yamada, 2000; Djurkovic et al, 2005; Lindy and Schaefer, 2010), lowered self-esteem (McMahon, 2000; Vickers, 2009) and even social isolation (Saam, 2010). In addition, bullies tend to erode a target’s professional reputation and may intend to unfavourably influence the target’s personal sense of ability (Vickers, 2009). In sum, bullying affects all facets of the target’s life (Townend, 2008). Specifically, job performance relates to assessment of performance at work (Sharma et al, 2009) and those who are bullied assess their work environment negatively (Stogstad et al, 2011). In such cases, the bullied individuals will perform less well at their jobs because of the bullying they experience. Earlier, some scholars (Yamada, 2000; Wu et al, 2011) have argued that bullying may result in reduced workplace productivity. Building on arguments presented in past literature, a negative impact of workplace bullying on job performance is expected.

Job Performance

As a job-related consequence, job performance concerns appraisal of the tasks that relate to an employee’s job and aims at achieving organisational goals (Sharma et al, 2009). A traditional perspective is that enhanced job performance will contribute to firm success. Thus, organisations should aim to maximise factors that are expected to boost job performance and eliminate factors that threaten it.

Emotional Intelligence

The EI construct has permeated modern academic as well as popular literature (Petrides and Furnham, 2001; Bar-On, 2005; Fernández-Berrocal and Extremera, 2006). Petrides and Furnham argue that the term appeared for the first time in an unpublished dissertation, yet it was Goleman’s (1998) work that popularised the construct. Since then, the construct has made steady theoretical advancement in the present century (Frazier et al, 2004; Bar-On, 2005), owing to claims that persons high in EI are more successful, healthier and happier (Carmeli, 2003; Ramo et al, 2009). In the domain of work performance, EI essentially represents an ability that precedes enhanced work performance (Boyatzis and Saatcioglu, 2008). For example, EI boosts creativity (Ivceic et al, 2007) through augmenting reason and providing flexibility (Isen, 1999), and regulates mood suitable for performance (Palfai and Salovey, 1993). It is also related with organisational commitment and a heightened sense of control at the workplace (Petrides and Furnham, 2006). Moreover, several investigations have confirmed the association between EI and extraordinary performance (Boyatzis, 1982; Humphrey, 2002; Wong and Law, 2002; Law et al, 2004; Sy et al, 2006). One might thus anticipate organisations to be specifically interested in the EI construct, as it may be developed and enhanced for individual and organisational gain (Jordan, 2005; Boyatzis and Saatcioglu, 2008), especially because people differ in their ability to focus on, manage and suitably apply emotions that pertain to themselves and others (Petrides and Furnham, 2006).

For this study, we use a conceptualisation of EI put forth by Petrides and Furnham (2001). They define trait EI as a set of self-perceived capabilities and behavioural dispositions evident in typical cross-situational behaviour that is, measurable through self-report. Several empirical studies have confirmed the utility of trait EI for desirable outcomes (Petrides et al, 2004; Petrides and Furnham, 2006; Mavroveli et al, 2007; Petrides et al, 2010), which makes it a valid and robust concept for the work domain.

How Emotional Intelligence May Moderate the Workplace Bullying/Job Performance Relationship

In interpersonal communications, emotionally intelligent persons are sensitive to emotions and feelings. Their response to a situation is not impulsive and hasty, but cautious and prudent. Thus, emotionally intelligent individuals make their emotions work for themselves by reflecting on and controlling actions to improve performance (Weisinger, 1988). Literature suggests that workplace bullying may trigger positive or negative emotions in one who is bullied, contingent upon how the target evaluates the bullying behaviour (Brotheridge and Lee, 2010).

We argue that one logic behind workplace bullying is that those who fall prey to it lack the EI to deal with the bully. Given that emotions play an important role in performance (Wagner and Ilies, 2008), it may be reasoned that individuals with high EI will not experience stress or threat when faced with a perpetrator, and will control their performance at work by regulating their emotions positively, not allowing bullying behaviour to influence their job performance, at least not in the short run. Given that bullying creates negative emotions in the target, such as mood changes, feelings of insecurity, anxiety, anger and even guilt (Ayoko et al, 2003; Namie and Namie, 2009), we argue that targets who are short on emotional resilience, that is the ability to fight back emotionally, are more predisposed to such negative emotions. By contrast, those high on emotional resilience and control are less likely to suffer adverse emotions. Thus, for those with high EI, detrimental effects on job performance due to bullying experienced at work are less likely, but may be anticipated among those low on EI.

Moreover, it is probable that the psychological damage caused by bullying (Fineman, 2004; Giorgi, 2010) can be remedied through appropriate emotional management, since emotionally intelligent individuals are adept at psychologically adjusting to adverse events (Fernández-Berrocal et al, 2006). Similarly, it may also be argued that the ability to repair oneself emotionally helps to overcome disturbing stressful thoughts (Salovey et al, 1995) evoked by bullying behaviour. It is this effective use of emotions that may conceivably moderate the damaging impact of bullying on work performance. Consequently, the impact of bullying on job performance will be less for those high on EI than for those low on EI. In sum, our framework posits that emotionally intelligent persons are better able to cope with bullying and thus weaken its negative consequences on their job performance.

On the basis of the arguments presented in the preceding section, the following relationships are proposed (Figure 1).

Figure 1
figure 1

Theoretical framework for the study.

Full size image

Research Hypotheses

This study will test the following hypotheses:

Hypothesis 1:

  • Workplace bullying negatively impacts on employee job performance.

Hypothesis 1a:

  • Belittlement negatively impacts on job performance.

Hypothesis 1b:

  • Work undermined negatively impacts on job performance.

Hypothesis 1c:

  • Verbal abuse negatively impacts on job performance.

Hypothesis 1d:

  • Workplace exclusion negatively impacts on job performance.

Hypothesis 2:

  • EI moderates the relationship between workplace bullying and job performance in such a way that bullying behaviour has lower impact on job performance for those high on EI and higher impact on job performance for those low on EI.

Methods and Materials

Sample

We employed a convenience sampling technique. A total of 400 questionnaires were distributed, and 242 usable replies were received, yielding a response rate of 60.5 per cent. The study sample comprised 132 men and 110 women. A total of 132 respondents were under 30 years of age, whereas 110 respondents were 30 years or older. As for work experience, 125 respondents had less than 5 years, while 117 respondents had 5 years or more.

Measures

Workplace bullying

We measured workplace bullying using Brotheridge and Lee’s (2010) instrument, which originally contained 43 items. To ensure cross-cultural conceptual similarity in measuring bullying, we followed guidelines suggested by Beaton and Guillemin (2000). Specifically, we sought the opinion of a committee comprising ten subjects from among the study respondents, and eight academics and subject-matter experts to confirm semantic and conceptual similarity of all workplace bullying items in the study context. Subsequently, discrepancies were revised and it was ensured that all items represented the main theme of the construct and corresponded to cultural settings. In this way, six items measured belittlement, five items work undermined, six items verbal abuse, and two items workplace exclusion. In all, 19 items measured workplace bullying.

Job performance

Job performance was measured using Blickle et al’s (2009) overall job-performance instrument that uses five items to assess task performance, adaptive performance and job dedication. In addition to these, we developed items to measure interpersonal facilitation and innovation aspects of job performance. Seven items measured total job performance.

Emotional intelligence

We used Petrides and Furnham’s (2001) trait EI instrument to measure EI. Ten items representing the main theme of the construct and covering the sub-scales were selected. A total of 36 positively worded items formed the research instrument for this study.

Data Collection

The survey questionnaire was floated personally, through e-mail, and through colleagues. A cover letter explained the purpose and academic nature of the study and assured the respondent of anonymity. Junior doctors were identified as trainees/medical students from Years 1 to 5 of medical studies, whereas senior doctors were identified as heads of departments. We distributed a total of 400 questionnaires to doctors working at five hospitals and six clinics in Islamabad and Rawalpindi and received 252 replies; ten contained a majority of blank responses and were dropped, resulting in 242 usable replies. The effective response rate was thus 60.5 per cent.

Analyses of Data and Results

Data were entered into SPSS 16.0 for analyses. Considering that researchers use control variables in cross-section research to test for differences within groups (Hardy, 1993), we used demographic variables of age, gender and years of work experience as control variables. Age was coded as ‘0’ for under 30, and ‘1’ for 30 or above; gender was coded as ‘0’ for male and ‘1’ for female; work experience was coded as ‘0’ for 5 or less years, and ‘1’ for 5 years or more.

We analysed demographic variables first, followed by descriptive statistics, reliability and correlation analysis. To confirm that the regression was non-spurious, we checked for regression assumptions of multicollinearity and homogeneity of variances of data before running the regression analysis. Multicollinearity indicates a strong relationship between two or more independent variables, and one way to test it is by examining the variance inflation factor and tolerance statistics. We found these values to be below 10 and 0.2, respectively, which ruled out multicollinearity (Field, 2005). Levene’s test was applied to test heteroscedasticity of variances, which was non-significant at p>0.05, indicating that variances were nearly equal. Following these tests, we tested the hypotheses and sub-hypotheses using forced-entry hierarchical regression (Tables 1 and 2).

Table 1 Demographic variables
Full size table
Table 2 Reliabilities, means, standard deviations and correlations among important variables
Full size table

Cronbach’s α was used to assess reliabilities of study variables. As Table 2 reveals, reliabilities for workplace bullying (α=0.87), belittlement (α=0.81), work undermined (α=0.72), verbal abuse (α=0.75), workplace exclusion (α=0.77), emotional intelligence (α=0.74) and job performance (α=0.90) are within the acceptable range, which indicates good internal consistency. Bivariate correlations were also examined between important variables. Workplace bullying and its dimensions positively correlated, as expected. It was also expected that job performance would correlate negatively with workplace bullying (r=−0.532). Job performance correlates negatively with workplace bullying dimensions of belittlement (r=−0.63), work undermined (r=−0.05), verbal abuse (r=−0.54) and workplace exclusion (r=−0.23). Meanwhile, EI correlates negatively with workplace bullying (r=−0.46) and its four dimensions of belittlement (r=−0.50), work undermined (r=−0.4), verbal abuse (r=−0.44) and workplace exclusion (r=−0.20). EI correlates positively with job performance (r=0.8).

Control Variables

To control for the effects of demographic variables of age, gender and years of work experience on job performance, we entered these in the first step of forced-entry multiple regression. An R2 value of 0.0071 and F statistic of 0.21 (p<0.76) showing the relation between control variables and job performance suggests that the controls had no significant relation with job performance.

Hypotheses Testing

Testing Hypothesis 1

Hypothesis 1 anticipated a negative relationship between workplace bullying and job performance. Sub-hypotheses 1a, 1b, 1c and 1d anticipated a negative relationship between job performance and the four dimensions of workplace bullying. We tested these using forced-entry ···· Table 3 presents the results. We tested this using forced-entry multiple regression. In the second step, we entered workplace bullying and its four dimensions of belittlement, work undermined, verbal abuse and workplace exclusion as predictors, and job performance as the criterion variable. Table 3 presents the results.

Table 3 Results of multiple regression analyses for Hypothesis 1, Hypothesis 1a, Hypothesis 1b, Hypothesis 1c, Hypothesis 1d
Full size table

As can be seen from Table 3, workplace bullying has a significant negative impact on job performance (β=−0.39, p<0.000). Hence, overall workplace bullying does have a significant negative impact on job performance. Observing the regression coefficients of workplace bullying dimensions, we see that belittlement (β=−0.67, p<0.000) has a significant negative impact on job performance, which confirms Hypothesis 1a. Unexpectedly, work undermined (β=0.09, p<0.000) has a weak and positive impact on job performance, which is contrary to Hypothesis 1b. Likewise, verbal abuse (β=0.13, p<0.000) has a significant and positive impact on job performance, which is contrary to Hypothesis 1c. Workplace exclusion (β=−0.34, p<0.000) was found to have a significant negative impact on job performance, which confirms Hypothesis 1d. Among the four workplace bullying dimensions, belittlement has the strongest impact on job performance. The overall model is significant (R2=0.45, F=20.8, p<0.000).

Testing Hypothesis 2

As mentioned earlier, we tested Hypothesis 2 using forced-entry hierarchical regression. To control for the effect of demographic variables (age, gender and years of work experience) on job performance, we entered these in the first step. In the second step, we added workplace bullying (predictor) and EI (hypothesised moderator) to see their impact on job performance. In the last step, an interaction term created as a product of standardised variables (Aiken and West, 1991) of workplace bullying and EI was added to examine the potential moderating effect of EI on workplace bullying and job performance. According to Baron and Kenny (1986), the moderation hypothesis is confirmed if the product term of the predictor and moderator (interaction term) is significant. The regression coefficient of the predictor is examined in the last step to test the strength and direction of the moderated effect. Change in R2 is also examined to see the additional change in the model attributable to the moderator (Frazier et al, 2004). We performed a simple slope test using unstandardised regression coefficients to plot regression lines, to see whether the slopes of the plotted lines differed from each other and from 0. Frazier et al (2004) recommend plotting a simple slope to see how the predictor and criterion relate to each other at different levels of the moderator (Table 4).

Table 4 Results of moderation analysis for Hypothesis 2
Full size table

Hypothesis 2 anticipated that EI would moderate the relation between workplace bullying and job performance. Following Baron and Kenny’s (1986) procedure to test the moderation effect, we added the predictor (workplace bullying) and hypothesised moderator (EI) in the second step, with job performance as the criterion variable. As expected, workplace bullying (β=−0.34, p<0.000) and EI (β=0.78, p<0.000) were significant in the regression model and the moderator appeared with a positive sign, as expected. The model is significant, because R2 is 0.83 and F statistic is 298.71 (p<0.000). Next, we tested for moderation in the last step by adding the interaction term in the model. The moderating effect of EI on the relationship between workplace bullying and job performance is confirmed, as their product term (interaction term) is significant (R2=0.841, F statistic=34.91, p<0.000). We examined the change in R2 to observe the additional change in the model attributable to the moderator (Frazier et al, 2004). The R2 change statistic indicates that EI affects an additional 0.041 or 4.1 per cent change in the model. These results provide support for Hypothesis 2. Figure 2 shows the interaction between workplace bullying and job performance at different levels of the moderator.

Figure 2
figure 2

Graphical representation of the workplace bullying and EI interaction for job performance.

Full size image

The preceding figure indicates that job performance is low for those with low levels and high for those with high levels of EI. In other words, the negative relationship between workplace bullying and job performance is weaker for those who are high on EI and stronger for those with low EI.

Discussion

The present study used data from 242 doctors in Islamabad and Rawalpindi with the objective of examining three propositions. One, whether workplace bullying in doctors negatively impacts on their job performance; two, which workplace bullying dimension impacts most on job performance; three, whether EI plays a moderating role in the relationship between workplace bullying and job performance, such that the negative impact of workplace bullying is lower for those who are high on EI and high for those who are low on EI. In all, of the four sub-hypotheses, the two main hypotheses (Hypothesis 1 and Hypothesis 2) and two sub-hypotheses (Hypothesis 1a and Hypothesis 1d) were confirmed, while findings conflicted with two sub-hypotheses (Hypothesis 1b and Hypothesis 1c). These results may be stated as follows.

This study found that workplace bullying indeed negatively impacts on job performance among doctors. Among dimensions of workplace bullying, belittlement had the greatest negative impact, followed by workplace exclusion. Unexpectedly, work undermined and verbal abuse did not impact on job performance negatively. Finally, this study also found that employee EI indeed moderates the relationship between workplace bullying and job performance, such that the performance of those who are high on EI is not as negatively impacted by bullying behaviour as performance of those who are low on EI, where it is more strongly negative. In other words, the interaction between workplace bullying and EI weakens the negative impact of bullying on job performance; although the relationship remains negative, it is more so for those low on EI and less so for those high on EI.

Theoretical Implications

Overall, results from this study support existing theory on the deleterious effects of workplace bullying (Djurkovic et al, 2005; Heames and Harvey, 2006; Carbo and Hughes, 2010) on work-related aspects. Results of this study clearly indicate that workplace bullying has adverse effects on employee job performance. Concerning dimensions of workplace bullying, belittlement had the strongest negative impact on job performance, followed by workplace exclusion.

Contrary to what was hypothesised, verbal abuse and work undermined did not exert negative impact, suggesting that when a supervisor verbally abuses employees or undermines an employee’s work, such bullying acts do not necessarily impede the employee's performance at work. Although this finding is somewhat surprising, it may be rationalised in the light of socio-cultural theory (Monks et al, 2009), according to which, bullying of any kind may be ingrained within organisational culture, particularly those of an authoritarian or hierarchical nature. As medical hospitals typically have a hierarchical organisational culture, verbal abuse and undermining may be perceived as acceptable, rather than as bullying behaviour. In particular, doctors encountering acts that involve verbal abuse and work undermining do not identify them as bullying behaviour, but as accepted components of that culture. Consequently, their job performance is not negatively impacted because they do not perceive themselves to have been bullied.

Alternatively, this finding may be rooted in a collectivist culture. Earlier, researchers have asserted that collectivist cultures differ from individualistic cultures in their coping strategy to deal with stressful events (Giorgi, 2010) and that people living in collectivist cultures turn to social support to deal with stressful events (Bhagat and Steers, 2009) such as workplace bullying. Giorgi (2010) also noted that people in collectivistic cultures are less likely to perceive negative acts as hurtful, compared with those living in individualistic cultures. Hence, when faced with such bullying behaviours, employees may turn to affiliations or group members to cope. Such social affiliations and memberships counteract the potential sense of oppression that verbal abuse and undermining are likely to provoke.

An interesting finding of this research is that the negative impact of workplace bullying is weaker for highly emotionally intelligent individuals, and stronger for individuals low on EI. These results highlight an important relevance of EI with regard to workplace bullying. Simply, highly emotionally intelligent individuals are capable of effectively coping with bullying at work. For them, EI is a coping mechanism, as it weakens the adverse effect of bullying on their job performance. Such individuals might direct their efforts towards maintaining their job performance, possibly to avert bullying.

Contribution to Literature

The present study confirms Boddy’s (2011) suggestion that bullies are likely to adversely affect employee performance. Furthermore, to the best of our knowledge, this study is the first to examine EI as a moderator to cope with workplace bullying for those who experience it. A novel contribution of this study is that it emphasises the role of both organisational and societal culture in the perception and outcomes related to workplace bullying. Certain bullying acts may be embedded within organisational culture and the target may not perceive them as such, thus cancelling the potentially damaging effect on job performance. Lastly, this study has examined the workplace bullying construct in Asian settings, where such research is lacking, as most studies related to bullying have been conducted in Western settings (Giorgi, 2010).

Managerial and Organisational Implications

Managers need to understand and define what bullying behaviour is and what it is not, specifically in the cultural and societal context where it is perpetrated so as to clarify specific behaviours that indicate bullying at work. This is because bullying behaviour may be subject to cultural sensitivities; a practical joke may imply bullying behaviour in the Western context, whereas the implication may be different in the Asian context. Clarification of this by managers and employees may distinguish when bullying has taken place and when it has not.

In sum, organisations must work towards removing bullying from the work environment. In this regard, we reassert the need for leadership to create an anti-bullying climate in organisations based on trust and integrity (Stogstad et al, 2011). An anti-bullying climate will lead to lower stress and anxiety levels among employees, higher commitment, reduced turnover, better employee health and consequently better job performance. Moreover, organisations may introduce anti-bullying policies to encourage an anti-bullying climate. Managers may also develop intervention strategies (Sheehan, 1999). Earlier studies (Overall, 1995) have suggested firm measures, such as legal recourse. Such remedial actions would ensure the emotional and physical health of employees; healthier, happier employees working in a bullying-free environment are expected to be more productive and their contribution to enhanced organisational performance would be an added benefit.

Workplace bullying is destructive in the emotional sense to the individual, and in the financial sense to the organisation (Carbo and Hughes, 2010; Boddy, 2011). Organisations aiming at improved performance must take into account the health and emotional aspects of their members. As prevention is better than cure, circumstances that lead to bullying can be avoided in a proactive manner. For one, organisations should consider whether an organisational change such as restructuring and downsizing, of which bullying is often a consequence (Sheehan, 1999), ought to take place or not.

Finally, attention must be paid to those who survive workplace bullying. Workloads increase for such employees as a result of global pressure and organisational change, in addition to bullying at work. Workplace counselling in such cases would be worthwhile to deal with adverse effects of bullying. Such managerial actions are likely to support a content, comfortable and emotionally healthy workforce that adds value to the organisation through superior performance.

Limitations

The cross-sectional nature of this study is a serious limitation. It is possible that the short-term effects of bullying may not impact on job performance for emotionally intelligent persons, but the results might change over time. After all, bullying is repetitive and insulting behaviour (Saam, 2010) and the emotional ‘strength’ of those who are targeted may be presumed to wear down over time. Another weakness of this study is that it has relied on employee perception of bullying behaviour, which overlooks whether such behaviour was intentional. It would be better, as discussed earlier, to have a common understanding of workplace bullying behaviour first. At the same time, since the nature of the bullying construct is based on the victim’s perception, it evidently excludes the intent of the alleged bully.

In addition, because this study uses data from doctors, the generalisability of its findings is questionable. A different sample from another industry such as education, engineering or telecommunications might yield different results. Further, as some scholars (Aguinis, 1995) have criticised the moderated multiple regression method, a limitation of this study may be its use of the product-term method to test moderation effects. Finally, this study has relied on self-reported data, which involves the risk of systematic errors.

Future Research Direction

As bullying behaviour may be directed towards an individual or group, as indicated earlier, we suggest group-level analysis in future research. Such a study would likely provide worthwhile findings for teams and groups in organisations. Moreover, future research may examine gender differences with regard to emotional responses within the studied relationship for a more holistic framework and theory-building. After all, women and men may differ in their reasons for going out to work, and this might affect how bullying behaviour affects their job performance.

As this research has proven that workplace bullying negatively impacts job performance, a similar adverse effect is likely on other aspects of work, such as turnover intentions, employee stress, workplace conflict, workplace climate and so on. Finally, we may expect bullying behaviour to be linked with certain personality traits such as neuroticism, as this trait is prone to anxiety and nervousness. Future research may examine this issue in depth.