Asia Pacific Journal of Human Resources (2019) , doi:10.1111/1744-7941.12234

Impact of abusive supervision on counterproductive work behaviors of nurses

Yoke May Low Universiti Putra Malaysia, Malaysia Murali Sambasivan Taylor’s University Lakeside Campus, Malaysia; Thiagarajar School of Management, India Jo Ann Ho Universiti Putra Malaysia, Malaysia

The objectives of this research are to address the two fundamental research questions: 1) What are the antecedents that lead to counterproductive work behavior (CWB) of nurses in public hospitals? 2) How effective are the moderating roles of power distance orientation (a cultural factor) and locus of control (an individual factor) in impacting CWB? The antecedents addressed are abusive super- vision and nurses’ perception of injustice. A questionnaire-based study was conducted among 337 nurses working in six public hospitals in different departments in Malaysia. The data were analyzed using structural equation modeling. The main findings are:1) nurses perceive injustice when sub- jected to abusive supervision and indulge in CWB; 2) nurses with a high level of power distance ori- entation are less likely to perceive abusive supervision as injustice; and 3) nurses with an internal locus of control are less likely to engage in CWB. The contributions of the study are discussed. Keywords: abusive supervision, counter-productive work behavior, perception of injustice, per- sonality factors

Key points 1 Abusive supervision does have impact on CWB. 2 Perception of injustice leads to CWB. 3 Abusive supervision directly and indirectly (through perception of injustice) leads to CWB. 4 Power distance orientation of nurses moderates the relationship between abusive supervision and perception of injustice. 5 Locus of control of nurses moderates the relationship between perception of injus- tice and CWB.

Correspondence: Murali Sambasivan, Faculty of Business and Law, Taylor’s University Lakeside Campus, Subang Jaya, Malaysia; e-mail: [email protected] Accepted for publication 1 May 2019.

© 2019 Australian HR Institute Asia Pacific Journal of Human Resources

Counterproductive work behavior (CWB) is not uncommon in the present, and it has sparked interest among many organizational researchers due to its pervasiveness over the past two decades (e.g. Berry, Ones and Sackett 2007; Burton and Hoobler 2011; Mitchell and Ambrose 2012; Penney and Spector 2002; Tepper 2000, 2007; Tepper et al. 2006; Wang et al. 2012). CWB is defined ‘as the volitional behaviors by employees that harm organizations and/or its members in the organization including peers, supervisors, cus- tomers and clients’ (Spector and Fox 2005, 151), and it emerges as one of the most severe problems faced by organizations in many nations (Chappell and Di Martino 2006). Researchers have found that 95% of employees have engaged in some form of CWB at least once (Penney and Spector 2002). Apart from the financial loss to businesses, numer- ous studies in the literature (e.g. Bennett and Robinson 2000; Bowling and Beehr 2006; Griffin and O’Leary-Kelly 2004; Ivancevich et al. 2003; LePine, Erez and Johnson 2002; Levine 2010; Podsakoff et al. 2009; Raman, Sambasivan and Kumar 2016) have suggested that the links people have with their supervisors, co-workers, and subordinates have a strong impact upon their personal well-being and success, including organizational mor- ale and productivity. Due to the disastrous effects of CWB, considerable research has been conducted to determine the antecedents of CWB (i.e. Bolin and Heatherly 2001; Marcus and Schuler 2004; Tepper 2007). One of the important antecedents of CWB is abusive supervision, which suggests that abused employees respond negatively to their supervisors by engaging in CWB (Inness, Barling and Turner 2005; Mitchell and Ambrose 2007; Tepper 2000). Abusive supervisors have been found to use their power and authority oppressively and vindictively towards their subordinates (Ashforth 1997). In addition, they ridicule and humiliate their subordinates publicly (Mitchell and Ambrose 2007). Such behavior can evoke negative reactions from the employees (Tepper 2007). Although abusive supervision has been given a great deal of attention, to the best of our knowledge there is a dearth of empirical research exploring how abusive supervision leads to CWB among nurses. Tepper (2000) has conducted research based on the justice model linking abusive supervision and organizational outcomes and has found that abu- sive supervision is a violation of supervisory justice. There are only a handful of studies that have examined the mediating effect of perception of injustice between abusive super- vision and CWB (e.g. Burton and Hoobler 2011; Wang et al. 2012). Nurses’ perceptions of injustice reflects the nurses’ perception of fair treatment on the job (Colquitt, Greenberg and Zapata-Phelan 2005). When abused employees experience abusive supervision, they believe that they have received lesser outcomes than they deserve in comparison to their target referents (Martin 1981). Abused nurses may find themselves in a less advantageous situation compared to their peers when their supervisors spend more time criticizing them than mentoring them for advancement. Equity theory and social exchange theory have emphasized the importance of social comparison in evaluating outcomes. For example, when someone receives favorable treatment, he/she will respond favorably (i.e. positive reciprocity), whereas one who receives unfavorable treatment will respond unfavorably

2 © 2019 Australian HR Institute Yoke May Low et al.

(i.e. negative reciprocity). Thus, CWB surfaces when an employee modifies his or her input to restore equity as a reaction to the perceived injustice of abusive supervision (Greenberg and Scott 1996). Therefore, when nurses are dissatisfied with the valuation of outcome fairness (e.g. being mistreated or abused), they will change their behavior to even the score and restore equity. Despite extensive research on the antecedents and consequences of CWB, very little research has investigated the contingencies affecting nurses’ perception of injustice and nurses’ decision to engage in CWB when subjected to abusive supervision. Prior researchers have shown that abusive supervision leads to CWB, but not everyone who experiences abusive supervision engages in CWB (Thau et al. 2009). The dissimilarities in values and characteristics among individuals may prompt them to act inversely towards abusive supervision. The process leading to nurses’ perception of injustice and CWB when subjected to abusive supervision is predicted to be determined by their own personalities and values. Accordingly, studies concerning abusive supervision and CWB with non-US samples have received attention from many international scholars (Martinko et al. 2013). However, to the best of the authors’ knowledge, these cultural values have yet to be studied in Malaysia, chiefly in the nursing context. The moderating role of power distance orientation and locus of control has been included in this study, as they are the most relevant cultural values in the current research context. The motivation to consider these constructs as moderators comes from recommendations by Tepper (2007) and Spector (2011). Since abusive supervision has been found to occur more commonly in high power distance countries, Tepper (2007) has called for greater concern for the impact of cultural values in studies regarding abusive supervision. Malaysia scores very high on the power distance dimension (score of 100) (‘Malaysia tops global power distance index’ 2014; Hofstede 2001). Hence, it has been predicted that power distance orientation moderates the relationship between abusive supervision and perception of injustice among nurses. Spector (2011) has claimed that it is wise and crucial to include variables of individual differences, such as locus of control in situations of injustice, for instance, as these constructs have been considered to be socially-learned and self-developed life behaviors. Hence, it has been predicted that locus of control moderates the relationship between abusive supervision and perception of injustice among nurses. Numerous studies found in the literature (e.g. LePine, Erez and Johnson 2002; Podsakoff et al. 2009) have suggested that the links people have with their supervisors, co- workers and subordinates exhibit a strong impact on their personal well-being and success, including organizational morale and productivity. Although studies pertaining to violence (e.g. Hockley 2002; Mayhew and Chappell 2001), aggression (e.g. Edward et al. 2014; Farrell, Bobrowski and Bobrowski 2006) and bullying (e.g. Hockley 2002; Hutchinson et al. 2006; Murray 2009) are abundantly available, only a handful of studies have looked into CWB in the nursing context. Additionally, most studies on CWB have primarily focused on nurse-to-nurse violence and aggression (e.g. Edward et al. 2014; Hockley 2002) instead of abusive supervision. Therefore, there is a need to study the impact of abusive supervision towards CWB among nurses.

© 2019 Australian HR Institute 3 Asia Pacific Journal of Human Resources

Prior studies have linked abusive supervision to CWB among nurses, whereby nurses who experience abusive supervision display a tendency to engage in CWB (Farrell, Bobrowski and Bobrowski 2006; Rosenstein and O’Daniel 2005). As a result, CWB dis- played by nurses as a form of retribution towards abusive supervisors may negatively affect the level of satisfaction among patients and jeopardize patient care, as nurses have the most contact with patients in hospitals (Farrell, Bobrowski and Bobrowski 2006; Rosenstein and O’Daniel 2005). Hence, it is essential to have a safe and supportive work- ing environment in the healthcare industry in order to increase the nursing workforce and to retain them within the healthcare system (Pillay 2017). The main contributions of this paper are twofold. First, we argue that abusive super- vision translates into CWB by nurses through the nurses’ perceptions of injustice. The empirical examination of nurses’ perception of injustice as a mediator may aid in compre- hending the underlying mechanism responsible for the relationship between abusive supervision and CWB among nurses. Therefore, this study offers further evidence con- cerning nurses’ perception of injustice as a mediator between abusive supervision and CWB. Second, we have studied the moderating roles of 1) power distance orientation between abusive supervision and nurses’ perception of injustice and 2) locus of control between nurses’ perception of injustice and CWB. Although mistreatment or injustice elicits retribution from those affected, not all abused employees engage in negative behav- iors. Hence, the model developed in this study discusses and argues for various individual differences (values and personalities) that could affect nurses’ decisions to engage in CWB. The impacts of power distance orientation and locus of control as moderators have not been considered in earlier studies related to nursing and the current study addresses this gap.

Literature review and hypotheses development

Counterproductive behaviors in the workplace have been investigated by many researchers under diverse headings, including workplace deviance (Martinko, Gund- lach and Douglas 2002), organizational delinquency (Hogan and Hogan 1989), organi- zation-motivated aggression (O’Leary-Kelly, Griffin and Glew 1996), bullying/ mobbing (Knorz and Zapf 1996), antisocial behaviour in organizations (Giacalone and Greenberg 1997), organizational retaliatory behaviour (Skarlicki, Folger and Tesluk 1999), workplace aggression (Fox and Spector 1999) and revenge (Bies and Tripp 1998). The diverse terminologies used are due to the varied theoretical perspectives under- taken by assorted researchers. For instance, Hogan and Hogan (1989) have based their work on criminological literature, and have compared incarcerated delinquents with a general college population in order to distinguish the characteristics of individuals who engage in acts they term organizational delinquency. In a similar manner, Bennett and Robinson (2000) have adopted the deviance approach, which involves violation of organi- zational norms and rules; O’Leary-Kelly, Griffin and Glew (1996) have based their work

4 © 2019 Australian HR Institute Yoke May Low et al. from the social psychological aggression literature; Skarlicki, Folger and Tesluk (1999) have employed the organizational justice approach in viewing CWB as a cognition-based response to experienced injustice; and Spector and colleagues (Fox and Spector 1999; Storms and Spector 1987) have portrayed CWB as an emotion-based response to stressful organizational environments. Bies and Tripp (1996) have proposed that certain organiza- tional events lead to negative emotions and cognition and eventually lead to vengeful acts, some of which could be CWB acts. These behavioral constructs describe similar aspects, despite being assigned different terms. According to Spector and Fox (2005), the common threads among the aforementioned research are: 1) these behaviors are executed by employees of an organization; 2) these behaviors are volitional as opposed to accidental; and 3) these behaviors harm or have the potential to harm an organization or its members. Many studies have been implemented to contribute to a better understanding of the essential antecedents of CWB (Hershcovis and Barling 2007; Marcus and Schuler 2004; Raman, Sambasivan and Kumar 2016). Researchers studying CWB have generally acknowledged individual difference variables as well as situational/organizational factors in determining the significant antecedents of CWB. For instance, an individual employ- ee’s characteristics are comprised of personal characteristics, which include but are not limited to affective traits, locus of control, narcissism, trait anger and trait anxiety, envy and the Big Five Personality traits (Mount, Ilies and Johnson 2006; Raman, Sambasivan and Kumar 2016). Organizational theorists favor research that portrays incidences and frequency of CWB in the light of situational factors (Robinson and Greenberg 1998; Spector and Fox 2010). These situational and organizational determinants include role conflict, role ambiguity, injustice, excessive workload, perceived control, interpersonal conflict, poor relations with peers and supervisors and abusive supervision (e.g. Tepper 2000, 2007). Despite a plethora of factors impacting CWB, the fundamental questions that remain to be addressed are: 1)What are the antecedents that lead to counterproductive work behavior (CWB) of nurses in public hospitals? and 2) How effective are the moder- ating roles of power distance orientation (a cultural factor) and locus of control (an indi- vidual factor) in impacting CWB? The purpose of this research is to enhance and to test a theoretical model of CWB associated with abusive supervision experienced at work, underscoring the psychological process and including the investigation of moderators. Although mistreatment or injustice elicits retribution from those affected, not all abused employees engage in negative behav- iors. Hence, the model developed in this study is discussed and argued from the perspec- tives of the various individual differences (values and personality) that could affect an individual’s decision to engage in CWB. Some researchers (e.g. Bies and Tripp 2001; Greenberg and Alge 1998) have argued that the motivation for revenge is often rooted in the perception of undeserved harm and injustice. However, it is individual differences that explain why certain employees, when subjected to the same poor treatment, engage in negative behaviors while some do not. The conceptual framework of this study is given in Figure 1.

© 2019 Australian HR Institute 5 Asia Pacific Journal of Human Resources

Nurses’ Abusive perception Counterproductive supervision of injustice work behaviors (AS) (NPI) (CWB)

Power distance Locus of orientation (PDO) control (LOC)

Figure 1 Research framework

Abusive supervision vs CWB Studies by Bies and Tripp (1996) and Bies (1999) suggest that threats made towards one’s identity can often lead to counterproductive responses. Furthermore, employees who perceive their supervisors to be a dominant source of interpersonal mistreatment (i.e. abusive supervision) that evokes negative emotions may desire retribution against the harm-doer (i.e. the supervisor) (Skarlicki, Folger and Tesluk 1999). Tepper (2000) has indicated that those abused hold their employers responsible for the behavior of their supervisors. Zellars, Tepper and Duffy (2002) have claimed that those abused withhold their organizational citizenship behaviors and display a higher tendency to engage in CWB (Detert et al. 2007). Moreover, Wei and Si (2013), who conducted research using 198 employees and their immediate supervisors from a multinational company in China, discovered that abusive supervision results in increased levels of CWB (i.e. sabotage, withdrawal, production deviance and theft). De Jonge and Peeters (2009) demonstrated that healthcare workers who are con- fronted with low emotional resources (e.g., emotional support from supervisors) display the tendency to exert CWB. In accordance with social exchange theory (Blau 1964), nurses who experience abuse are likely to rectify such poor treatment by engaging in negative behaviors in order to even the score. One can expect that supervisory mistreatment encourages retaliatory behavior (Mitchell and Ambrose 2007; Skarlicki and Folger 1997). So, when the nurses’ supervisors fail to treat them with respect or with justice, nurses then seek to vent their frustration by acting counterproductively. Hence, consistent with prior research, nurses abused by their supervisors are believed to have more tendencies to exert CWB. Thus, it is hypothesized that:

Hypothesis 1: Abusive supervision has a positive relationship with nurses’ CWB.

Mediation effect of nurses’ perception of injustice between abusive supervision and CWB When one suffers from abusive supervision, an intervening psychological process leads to behavioral reactions. The studies carried out by Martinko, Gundlach and Douglas (2002) and Spector and Fox (2002) have shown that targets have diverse psychological reactions (e.g. negative mood and perceptions of injustice) after being exposed to aversive stimuli.

6 © 2019 Australian HR Institute Yoke May Low et al.

The association between abusive supervision and CWB is mediated by a cognitive psycho- logical process termed ‘nurses’ perception of injustice’. Generally, subordinates observe injustice by comparing the fairness between their inputs to the organization and the out- puts they gain from the organization. Past investigations have suggested that perceived injustice is linked to CWB. For instance, Greenberg and Scott (1996) have confirmed that counterproductive and withdrawal behaviors surface when an employee modifies his or her input to restore equity as a reaction to perceived injustice. This study proposes that when nurses perceive that their supervisors are being unfair in resource allocation, they tend to harm their supervisors. Therefore, when these nurses are dissatisfied with the valuation of outcome fairness (e.g. being mistreated or abused), they will change their behavior to even the score and to restore equity. Perceived injustice can eventually lead to resentment towards offenders and victims (i.e. CWB) (Ambrose, Seabright and Schminke 2002; Skarlicki and Folger 1997). Hence, in agreement with equity theory (Adams 1965) and other prior research, perception of injustice among nurses leads to CWB. When nurses perceive injustice, they perceive that their supervisors have failed to treat them with respect and to fulfill their personal needs at work. Thus, it is hypothesized that:

Hypothesis 2: Nurses’ perception of injustice mediates the relationship between abusive treat- ment by the supervisor and the nurses’ inclination to engage in CWB.

Moderating role of power distance orientation between abusive supervision and nurses’ perception of injustice Power distance is related to social inequality, as well as the amount of authority a person has over others (Hofstede and Bond 1984). In other words, power distance refers to the acceptance of unequal dispersal among members of institutions and organizations (Hofstede 1980). Although Hofstede discovered that studies regarding cultural values are meaningful only at the societal level, Kirkman and Shapiro (2001), as well as Clugston, Howell and Dorfman (2000), have noted huge variation among individuals in societies in each of Hofstede’s value dimensions, and these individual differences directly affect many outcomes. Power distance at the individual level refers to ‘the extent to which an individual accepts unequal power dissemination in institutions and organizations’ (Clugston, Howell and Dorfman 2000, 9). This variable of individual distinctions shapes a person’s relationship with his or her authority figures. Managers from high power distance countries are more task-oriented and less people-oriented, as they see the role of a manager in a high power distance system as solely to initiate structure (Hofstede 1980). According to Hofstede (1980), individuals from high power distance countries behave submissively in the presence of managers. They feel intimidated or at least unwilling to go against their superiors. They assent to inequality in power distribution and consider it normal in their society (Nicholson and Stepina 1998). Hence, individuals with a high power distance orientation are less likely to form links with their superiors, in comparison to those with a low power distance orientation (Begley et al. 2002). This is because

© 2019 Australian HR Institute 7 Asia Pacific Journal of Human Resources individuals with a higher power- distance orientation believe that they have less signifi- cance in influencing the decision-making process made by their superiors, whereas those with a low power distance orientation are expected to be involved in the decision-making process (Begley et al. 2002; Lam, Schaubroeck and Aryee 2002). Lee, Pillutla and Law (2000) and Lam, Schaubroeck and Aryee (2002) have asserted that subordinates with a high power distance orientation are able to accept arbitrary treatment. Thus, they are less likely to question the authority of superiors, mainly because they believe in power inequal- ity between superiors and subordinates. Nurses in a high power distance orientation are less sensitive to abuse and therefore do not view abuse as abusive, in comparison to their lower power distance counterparts. This may be due to their assumption that the behavior of supervisors is not harmful, as they hold certain implicit beliefs regarding unequal power distribution. Hence, they are less likely to perceive abusive supervision as unjust and are less likely to reciprocate accordingly. They further believe that their supervisors possess higher power, and thus are afraid or at least unwilling to disagree with their supervisors. In contrast, nurses with lower power distance values are expected to be less likely to accept arbitrary treatment from their supervisors, and hence will reciprocate by engaging in CWB because they have the tendency to perceive abusive supervision as unfair. Thus, it has been hypothesized that:

Hypothesis 3: Power distance orientation moderates the relationship between abusive supervi- sion and nurses’ perception of injustice; nurses who have higher power distance are less likely to perceive injustice when subjected to abusive supervision compared to those with low power distance.

Moderating role of locus of control between nurses’ perception of injustice and CWB Locus of control refers to one’s general beliefs about the controllability of events in one’s life (Spector 1988). This personality factor represents an individual’s beliefs concerning his or her control over the environment. Moreover, Rotter (1966) defines locus of control as an individual’s generalized expectancy or belief pertaining to the nature of outcomes in his or her life. Those who believe that they can control whatever happens to them (per- sonal control) are said to possess an internal locus of control (self-control) (Rotter 1966). Since they believe that they can control their lives, they accept responsibility for the events that occur (Davis and Davis 1972). Conversely, individuals with an external locus of con- trol orientation believe that results are generally beyond personal control and that their fate is controlled by other powerful external forces, such as luck, chance and fate (Rotter 1966). Individuals with this personality type tend to blame their surrounding for failures (Phares, Wilson and Klyver 1971). Individuals who perceive that they have lower control engage in acts of destruction under certain conditions. The act of destroying objects or parts of the physical environ- ment increases an individual’s feeling of control. Researchers have determined that indi- viduals with an external locus of control are more likely to use maladaptive coping

8 © 2019 Australian HR Institute Yoke May Low et al. strategies. People with an internal locus of control are more resistant to social pressure, and are less willing to hurt people, even when ordered to do so by an authority figure (Quick and Nelson 2009). They are also more likely to act upon that which they think is right and be responsible for their actions (Crooker, Smith and Tabak 2002). The link between nurses’ perception of injustice and CWB is predicted to be stronger when the abused nurses have an external locus of control. Moreover, nurses with an external locus of control believe that outcomes generally exceed their personal control and that their fate is being controlled by other powerful external forces, such as luck. Accordingly, they tend to blame their environment for failure. In this case, the frustrated external locus of control nurses will automatically blame their supervisors for abusing them, and hence are more likely to perceive injustice. In addition, as locus of control predicts sensitivity to injustice, nurses with an external locus of control may attempt to modify their environment solely to increase their feeling of control (Quick and Nelson 2009). Therefore, they defend themselves in a neg- ative way by engaging in CWB as they perceive a lack of control over events and their own inability to hinder unfortunate events. Thus, has been hypothesized that:

Hypothesis 4: Locus of control moderates the relationship between nurses’ perception of injustice and CWB; nurses with an internal locus of control are less likely to react to perceived injustices by resorting to CWB compared to those with an external locus of control.

Methods

Population, sampling and sample size The research population incorporates nurses from hospitals in Malaysia. There are 153 public and 216 private hospitals in Malaysia. According to the Ministry of Health (MOH), the number of beds in the public and private hospitals is 61 299 and public hospitals account for 69% of beds. Klang Valley in Malaysia accounts for nearly 25% of the coun- try’s population and major public hospitals are located in this area. Therefore, six large public hospitals in Klang Valley were chosen for this study. Nurses from public hospitals are state registered nurses (SRNs) with grade U29 and above, holding a valid Malaysian Nursing Board License that allows them to practice nursing in public hospitals, with the minimum qualification of a certificate or diploma in nursing. There are approximately 94 000 nurses in Malaysia, as reported by the MOH in health facts 2016. SRNs with grade U29 and above handle patients and report directly to a nurse supervisor. Based on the table provided by Krejcie and Morgan (1970), the required sample size is at least n = 384. Furthermore, since this study employed a self-administered questionnaire for data collection, there was a concern regarding low response rate. Therefore, the sample size was inflated by 50% (Bryman and Bell 2007), resulting in a total of 576. This figure was fur- ther rounded up, and the sample size of 580 nurses was finally decided upon for this study. Six hospitals from the Klang Valley were chosen due to the higher density of popula- tion with more well-equipped hospitals concentrated in this area. The six hospitals selected for this study have a total capacity of 5632 beds. The number of questionnaires

© 2019 Australian HR Institute 9 Asia Pacific Journal of Human Resources distributed was based on the proportion of beds at each hospital. The data collection was done after obtaining the necessary permission from the MOH. The questionnaires were given to the head nurse at each hospital and the head nurse distributed them to the nurses at various departments. A pilot study was conducted with 35 nurses from different hospi- tals. The pilot study indicated that the sampled nurses did not raise any concerns about the items in the questionnaire.

Measures Abusive supervision was measured by using Tepper’s (2000) 15-item abusive supervision measure to assess perceptions given by respondents towards abusive behaviors exerted by their supervisors. For example, some of the items included in this construct are: ‘supervisor tells me my thoughts or feelings are stupid’; ‘ridicules me’; and ‘invades my privacy’. In order to gather data, the respondents were asked to indicate the frequency of experiencing abusive behaviors from their supervisors with each item, using a five-point Likert scale labelled ‘never’ (1), ‘less than 5 times a year’ (2), ‘more than 5 times a year’ (3), ‘once a month’ (4), and ‘once a week or more’ (5). This construct was treated as a formative construct (Tepper 2000). Nurses’ perception of injustice was measured by using Colquitt’s (2001) 20-item scale. It was adapted in this study to measure the perceptions of injustice among subordinates about their current job. This construct had three dimensions: procedural injustice (seven items), distributive injustice (four items) and interactional injustice (nine items). For example, some of the items included in this construct are: 1) ‘Have you been able to express your views and feelings during those procedures?’ 2) ‘Does your outcome reflect the effort you have put into your work?’ and 3) ‘Has he/she treated you in a polite man- ner?’ This measure asks the respondents to determine their agreement or otherwise with each item in relation to their current job on a five-point Likert scale labelled ‘strongly dis- agree’ (1), ‘disagree’ (2), ‘neutral’ (3), ‘agree’ (4), and ‘strongly agree’ (5). This construct was treated as a reflective-formative higher order construct (Colquitt 2001). CWB was measured by using a 24-item measure adapted from Mitchell and Ambrose (2007) and Bennett and Robinson (2000). The CWB construct consisted of three dimen- sions: supervisor-directed (10 items), customer-directed (four items) and organization- directed (10 items). For example, some of the items included in this construct are: 1) ‘made fun of my supervisor’, 2) ‘argued or fought with a customer’ and 3) ‘called in sick when not’. These measures, on a five-point Likert scale, had the respondents indicate the frequency of engagement in the behaviour described in the previous year (targeting their supervisor, customers and the organization for which they had been working), ranging from ‘never’ (1), ‘once a year’ (2), ‘several times a year’ (3), ‘weekly’ (4), and ‘daily’ (5). This construct was treated as a formative-formative higher order construct (Mitchell and Ambrose 2007). Power distance orientation was measured by using a six-item scale developed by Dorfman and Howell (1988) and Farh, Hackett and Liang (2007). For example, some of the items included in this construct are: 1) ‘Managers should make most decisions

10 © 2019 Australian HR Institute Yoke May Low et al. without consulting subordinates’; and 2) ‘Employees should not disagree with manage- ment decisions. The respondents were asked to indicate their response for each of the behaviours using a five-point Likert scale, labelled ‘strongly disagree’ (1), ‘disagree’ (2), ‘neutral’ (3), ‘agree’ (4), and ‘strongly agree’ (5). This construct was treated as a reflective construct (Farh, Hackett and Liang 2007). Locus of control was measured using a 16-item Work Locus of Control Scale (WLCS) adopted from Spector (1988). According to Spector (1988, 1), locus of control is defined as a generalized expectancy that rewards, reinforcements or outcomes in life are controlled either by one’s own actions (internal locus of control) or by other forces (external locus of control). For example, some of the items included in this construct are: 1) ‘A job is what you make of it, 2) ‘Making money is primarily a matter of good luck’ and 3) ‘People who perform their jobs well generally get rewarded’. The respondents were asked to indicate their beliefs about their job in general using a five-point Likert scale, labelled ‘strongly dis- agree’ (1), ‘disagree’ (2), ‘neutral’ (3), ‘agree’ (4), and ‘strongly agree’ (5). The low mean score represents an internal locus of control and a high mean score represents an external locus of control. This construct was treated as a formative construct (Spector 1988).

Questionnaire response rate Out of the 580 distributed questionnaires, a total of 337 responses were deemed usable, thus yielding a response rate of 58.10%.

Reliability of constructs A reliability test is essential in determining the goodness of the data gathered. Cronbach’s Alpha (a) is the reliability coefficient that indicates the positive correlation of one item to another in a set (Sekaran 2000). The closer the value of Cronbach’s alpha (a) to 1.000, the higher the reliability of the research instruments (Sekaran 2000). In addition, Sekaran (2003) claims that value of reliability with Cronbach’s alpha less than 0.600 is considered poor, while those in the range of 0.700 are acceptable, and those over 0.800 are considered good. The alpha reliability for abusive supervision, nurses’ perception of injustice, CWB, power distance orientation and locus of control were 0.870, 0.920, 0.810, 0.810 and 0.850, respectively.

Handling common method bias Since the data were captured from one source (nurses), common method variance (CMV) was carried out using Harman’s Single Factor technique in order to capture any common method bias present in the data set (Jarvis, MacKenzie and Podsakoff 2003; Podsakoff et al. 2003). Initially, all the measurement items were entered to perform a factor analysis using SPSS software program. The results obtained showed that the largest variance explained by the first factor is 10.33% of the total variance. Furthermore, the factor analysis did not display any new general factor from the data set. Hence, it is suggested that the common method bias is insignificant in the data set obtained for this study.

© 2019 Australian HR Institute 11 Asia Pacific Journal of Human Resources

Measurement and structural model analyses tool According to Hair et al. (2014), structural equation modeling (SEM) is a second genera- tion multivariate data analysis method. There are two approaches available in estimating the correlations in SEM; covariance-based SEM (CB-SEM) and variance-based SEM (PLS-SEM). In the context of this research, we used Partial Least Squares-Structural Equation Modeling (PLS-SEM) 3.0 to run the measurement and structural models for several reasons. First, the purpose of this research is exploratory. We seek to explain why abused nurses engage in CWB. Abusive supervision has been examined to determine if it exerted any effect upon nurses’ perception of injustice and CWB. In addition, this study tests if power distance orientation has any influence on the relationship between abusive supervision and nurses’ perception of injustice. The effect of locus of control has been examined in the relationship between nurses’ perception of injustice and CWB. Second, the structural model of this research consists of four formative measured constructs. Abu- sive supervision and locus of control are formative measured constructs. On the other hand, nurses’ perception of injustice has been measured in a reflective-formative higher order construct, while CWB has been measured as a formative-formative higher order construct. According to Becker, Klein and Wetzels (2012) and Hair et al. (2014), PLS- SEM is the preferred tool when a structural model has formative and reflective constructs. The terms formative and reflective are used as a method of analysis by PLS-SEM users. A formative measurement model (also known as composite factor model) is a type of measurement model setup in which the direction of the arrows is from the indicator variable to the construct, indicating the assumption that the indicator variables cause the measurement of the construct (Hair et al. 2014). A reflective measurement model (also known as common factor model) on the other hand is a type of measurement model setup in which the direction of the arrows is from the construct to the indicator variables, indicating the assumption that the construct causes the measurement (covariation) of the indicator variables (Hair et al. 2014). Reflective-formative constructs and formative- formative constructs refer to the use of hierarchical latent variable models of PLS-SEM in more advanced and complex models (Becker, Klein and Wetzels 2012).

Results

Profile of the respondents A total of 337 nurses returned the questionnaire (response rate = 58%). About 94% of the respondents were female, and 95% of the respondents were less than 39 years of age. About 82% of the nurses had fewer than 9 years of experience. The nurses came from different departments such as outpatient, surgery, pediatrics, orthopedics, emergency, psychiatry, and gynecology.

Descriptive statistics and correlation A useful statistic that emerged from this study is the list of activities that constitute abusive supervision and CWB in Malaysian public hospitals. Table 1 provides a list of the top 10

12 © 2019 Australian HR Institute Yoke May Low et al.

Table 1 Top 10 cases of abusive supervision and CWB in Malaysian Public Hospitals Abusive supervision incidents % CWB incidents %

Gives me the silent treatment. 59.30 Swore at my supervisor. 54.90 Does not give me credit for jobs 56.60 Called in sick when not. 54.00 requiring a lot of effort. Ridicules me. 56.60 Endangered customers by not 50.70 following safety procedures. Does not allow me to interact with 56.40 Argued or fought with customers. 48.40 my peers. Invades my privacy. 52.00 Came in late to work without 47.50 permission. Is rude to me. 52.00 Physically attacked a customer 46.90 (pushing/hitting). Makes negative comments about 50.70 Made an ethnic, religious or racial 46.90 me to others. remark against my supervisor. Lies to me. 50.10 Publicly embarrassed my 46.60 supervisor. Break promises he/she makes. 50.00 Littered my work environment. 44.20 Tells me my thoughts or feelings are 36.80 Made an obscene comment or 38.60 stupid. gesture towards my supervisor.

Table 2 Descriptive statistics for all constructs (N = 337) Construct Sub-dimensions Reliability Mean (min/max) Std. deviation

Abusive Supervision .870 3.00 (2.00/3.93) .40 Nurses’ Perception Nurses’ Perception of .920 3.26 (2.00/4.57) .42 of Injustice Procedural Injustice Nurses’ Perception of 3.50 (1.50/5.00) .57 Distributive Injustice Nurses’ Perception of 3.42 (2.00/5.00) .53 Interactional Injustice Overall 3.40 (2.00/4.45) .39 Counterproductive CWB targeted at .810 2.47 (1.60/3.50) .38 Work Behaviors Supervisor CWB targeted at 2.63 (1.60/3.50) .33 Organization CWB targeted at 3.14 (1.75/4.25) .51 Patients Overall 2.65 (1.75/3.42) .28 Power distance .810 2.56 (1.00/4.67) 1.02 orientation Locus of control .850 2.52 (1.44/3.75) .43

© 2019 Australian HR Institute 13 Asia Pacific Journal of Human Resources

Table 3 Correlations between Abusive Supervision, Nurses’ Perception of Injustice, Counterpro- ductive Work Behaviors (CWB), Power Distance Orientation, and Locus of Control Constructs AS NPI CWB PDO LOC

AS 1 NPI .298* 1 CWB .338** .454** 1 PDO .161** .175** .171** 1 LOC .003 .105* .009 .020 1 **p < 0.010, *p < 0.050. AS = abusive supervision; CWB = counterproductive work behaviors; LOC = locus of control; NPI = nurses’ perception of injustice; PDO = power distance orientation. activities and the percentage of nurses experiencing them. Table 2 provides the mean, minimum and maximum scores, standard deviation, skewness score, and kurtosis score for each construct used in the study. The mean score for abusive supervision (3.00), with a minimum score of 2.00 and a maximum score of 3.93, indicates that the nurses perceive that their supervisors have engaged in abusive behaviors towards them (e.g. ridiculing them, silent treatment, invading their privacy) at least five times annually on an average. The mean scores for the different dimensions of nurses’ perception of injustice and an overall mean score of 3.40, with a minimum score of 2.00 and a maximum score of 4.45, indicate a moderate level of perception of injustice at the workplace. The mean scores for dimensions of CWB and an overall mean score of 2.65, with a minimum score of 1.75 and a maximum score of 3.42, indicate that the nurses have engaged in CWB, few times a year. A low mean score on the locus of control (2.52), with a minimum score of 1.00 and a maximum score of 4.67, indicate that the locus of control of nurses varies between internal and external. A mean score of 2.56, with a minimum score of 1.44 and a maximum score of 3.75, indicate the existence of low to moderate levels of power distance orientation in public hospitals in Malaysia. Table 3 provides the correlation between the constructs.

Reliability and validity (based on the measurement model) In the framework, there are two reflective constructs: nurses’ perception of injustice and power distance orientation. Based on the measurement model, the indicators that had factor loading less than 0.708 (Hair et al. 2014) were removed from further analysis. Accordingly, two items were removed from nurses’ perception of injustice (out of a total of 20 items) and power distance orientation (out of a total of six items). As shown in Table 4, the composite reliability (CR) and average variance extracted (AVE) of these two constructs are 1) 0.903 and 0.547 (nurses’ perception of injustice) and 2) 0.979 and 0.920 (power distance orien- tation). These values meet the guidelines recommended by Hair et al. (2014) (≥0.700 for CR and ≥0.500 for AVE). For checking the discriminant validity, Heterotrait-Monotrait Ratio of Correlations (HTMT) criteria was used. From Table 5, the HTMT score of 0.219 is within the range (1 < HTMT < 1) recommended by Henseler, Ringle and Sarstedt (2015).

14 © 2019 Australian HR Institute Yoke May Low et al.

Table 4 Summary of results for reflective constructs Construct Indicators Scale Indicator Composite AVE Convergent Discriminant reliability reliability validity validity (loadings) (AVE > 0.5)

Nurses NPJ3 Reflective .622 0.903 0.547 Yes Yes Perception NPJ4 .829 of Injustice NPJ5 .763 (NPI) NPJ6 .755 NPJ7 .539 NDJ1 .761 NDJ2 .760 NDJ3 .819 NDJ4 .798 NIJ1 .529 NIJ2 .591 NIJ3 .794 NIJ4 .613 NIJ5 .719 NIJ6 .839 NIJ7 .873 NIJ8 .813 NIJ9 .767 Power PDO1 Reflective .952 .979 .920 Yes Yes Distance PDO2 .964 Orientation PDO3 .969 (PDO) PDO5 .950 Indicator Reliability 0.700 or higher is preferred. If exploratory research, 0.400 or higher is accept- able (Hulland, 1999).

Table 5 Heterotrait-Monotrait Ratio of Correlations (HTMT) Criterion Analysis for discriminant validity NPI PDO

NPI PDO .219

HTMTinference (1.000 < HTMT < 1.000) (Henseler, Ringle and Sarstedt 2015). NPI = nurses’ perception of injustice; PDO = power distance orientation.

According to Bollen (2011), validity makes little sense for formative indicators. There- fore, all the formative indicators were retained for further analysis, as omitting formative indicators may to some extent omit some of the constructs’ content (Hair et al. 2014).

© 2019 Australian HR Institute 15 Asia Pacific Journal of Human Resources

Table 6 Formative constructs significance testing results Constructs Indicators Scale Weights VIF t-value Weights

AS AS1 Formative .072 1.602 .442 AS2 .083 1.472 .643 AS3 .415 1.119 3.948** AS4 .271 1.678 1.941 AS5 .025 1.319 .061 AS6 .044 1.435 .306 AS7 .158 2.176 .964 AS8 .097 1.955 .614 AS9 .542 1.973 3.731** AS10 .157 1.966 1.004 AS11 .457 1.592 3.790** AS12 .239 1.880 1.540 AS13 .057 1.965 .326 AS14 .211 1.753 1.463 AS15 .320 1.712 2.134* LOC LOC1 Formative .693 2.011 2.225* LOC2 .723 2.048 2.696** LOC3 .296 1.803 1.592 LOC4 .182 1.659 .253 LOC5 .566 4.629 .368 LOC6 .088 1.268 1.047 LOC7 .122 1.521 1.037 LOC8 .243 10.157 DELETED LOC9 .065 4.515 .187 LOC10 .544 4.048 .097 LOC11 .246 1.606 1.232 LOC12 .248 1.359 2.194* LOC13 .308 4.232 1.016 LOC14 .145 1.593 .610 LOC15 .064 1.404 .470 LOC16 .775 7.915 DELETED NPI NPJ Second Order Formative .262 1.309 7.859** NDJ .280 1.556 14.051** NIJ .691 1.408 22.584** CWB CWBS1 Formative .023 1.304 1.042 CWBS2 .114 1.197 .208 CWBS3 .451 1.613 1.516 CWBS4 .292 1.798 2.710** CWBS5 .047 1.266 .788 CWBS6 .026 1.166 1.379 CWBS7 .275 1.262 1.320

16 © 2019 Australian HR Institute Yoke May Low et al.

Table 6 (continued) Constructs Indicators Scale Weights VIF t-value Weights

CWBS8 .359 1.959 3.230** CWBS9 .185 1.733 1.514 CWBS10 .076 1.200 .132 CWBO1 .348 1.916 3.054** CWBO2 .365 1.627 2.517* CWBO3 .270 1.661 2.436* CWBO4 .259 1.417 2.124* CWBO5 .319 1.788 2.001* CWBO6 .077 1.835 .676 CWBO7 .774 7.827 DELETED CWBO8 .186 1.223 .939 CWBO9 .067 1.176 .515 CWBO10 .038 1.314 .107 CWBP1 .263 1.228 3.459** CWBP2 .385 1.947 2.653** CWBP3 .125 2.372 1.857 CWBP4 .586 1.771 4.487** VIF ≤ 5.000 (Hair et al., 2014), **p < 0.010, *p < 0.050. AS = abusive supervision; CWB = counterproductive work behaviors; LOC = locus of control; NPI = nurses’ perception of injustice.

Table 6 provides information on the outer weights and VIF for each item of formative constructs. As indicated in the table, there were three items (two items of locus of control and one item of CWB) that had VIF greater than the recommended value (VIF > 5.000) and they were removed (Hair et al. 2014).

Structural model results We followed the five-step procedure recommended by Hair et al. (2014). The five steps involve assessment of 1) collinearity issues, 2) path coefficients, 3) R2 values, 4) effect size and 5) predictive relevance, Q2. Table 7 provides information on the collinearity assess- ment. Since the VIF < 5.000, there is no collinearity problem in this model. Tables 8 and 9 give the path coefficients of all the paths and it can be seen that all the path coefficients are significant. Table 10 indicates the R2 values for nurses’ perception of injustice (R2 = 0.148 – moderate) and CWB (R2 = 0.310 – substantial). The effect sizes and their interpretation are given in Table 10. The assessment of predictive relevance Q2 accurately forecasts the relative predictive relevance of an exogenous construct on an endogenous construct (Hair et al. 2014). As well, Hair et al. (2014) have recommended that the rule of thumb be that for Q2 values greater than zero for a particular reflective endogenous latent variable. However, this

© 2019 Australian HR Institute 17 Asia Pacific Journal of Human Resources

Table 7 Collinearity Assessment AS NPI CWB

AS 1.224 1.000 NPI 1.475 CWB VIF ≤ 5.000 (Hair et al., 2014). AS = abusive supervision, CWB = counterproductive work behaviors; NPI = nurses’ perception of injustice.

Table 8 Path co-efficient assessment (N = 337) Hypothesis Relationship Direct effect, b Standard deviation (STDEV) t-Value p-Value

H1 AS ? CWB .167 .071 2.364* .009

H2 AS ? NPI .385 .047 8.235** .000

H3 NPI ? CWB .471 .057 8.222** .000 **p < 0.010, *p < 0.050. AS = abusive supervision; CWB = counterproductive work behaviors; NPI = nurses’ perception of injustice.

Table 9 Mediation result Hypothesis Relationship Beta Bootstrapped Confidence Interval

Indirect Standard t-Value p-Value Result 95% 95% effect a*b error LL UL

H4 AS ? NPI .181 .035 5.232** .000 Significant .136 ? CWB .284 **p < 0.010, *p < 0.050. AS = abusive supervision; CWB = counterproductive work behaviors; NPI = nurses’ perception of injustice. assessment is not applicable in this study, as all the endogenous constructs are formative in nature (Hair et al. 2014). Hence, the procedure is omitted in this study. The mediation test was carried out using the approach introduced by Preacher and Hayes (2008), which incorporated bootstrapping the sampling distribution of the indirect effect. This approach is most relevant for PLS-SEM method because bootstrapping does not make any assumption regarding the sampling distribution of the statistics, and more- over it is applicable to smaller sample sizes with higher confidence (Hair et al. 2014).The result indicates that nurses’ perception of injustice mediates the relationship between abu- sive supervision and CWB (b = 0.181, t-value = 5.232, p-value = 0.000). Hair et al.

18 © 2019 Australian HR Institute Yoke May Low et al.

Table 10 The determination of co-efficient (R2) and effect size (f2)(N = 337) Co-efficient of determination Effect size (f2)

R2 NPI Effect size CWB Effect size

AS – .006 Small effect .034 Small effect NPI .148 – .274 Medium effect CWB .310 –– Note. 1) Effect size f2 interpretation: Cohen (1988) suggested 0.020 as small effect, 0.150 as medium effect, and 0.350 as large effect (Hair et al. 2014); 2) R2 values interpretation: Cohen (1988) sug- gested 0.020 as weak, 0.130 as moderate, and 0.260 as substantial. AS = abusive supervision; CWB = counterproductive work behaviors; NPI = nurses’ perception of injustice.

(2014) have suggested that the extent to which the variance of CWB is directly explained by abusive supervision and the variance of the target construct can be explained via indirect relationship, where nurses’ perception of injustice can be assessed by using vari- ance accounted for (VAF). As such, VAF can be calculated using the following formulae: VAF ¼ða bÞ=c where a*b indirect effect; c total effect (indirect effect + direct effect). In this framework, direct effect indicates the direct relationship between abusive supervision and CWB without any intervention; indirect effect indicates the relationship between abusive supervision and CWB through nurses’ perception of injustice (mediating effect). The calculated VAF score for Hypothesis is 52%, signifying partial mediation (Hair et al. 2014). Hence, the relationship between abusive supervision and CWB had been par- tially mediated by nurses’ perception of injustice. The moderating effect of power distance orientation (Hypothesis 3) between abusive supervision and nurses’ perception of injustice is significant (b = 0.319, t-value = 3.475, p-value = 0.000). Figure 2 shows the moderation relationship. The results show that nurses with a higher level of power distance orientation are less likely to perceive abusive super- vision as injustice. On the contrary, nurses are more likely to perceive injustice when sub- jected to abusive supervision when there is a low level of power distance orientation. The moderating effect of locus of control (Hypothesis 4) between nurses’ perception of injustice and CWB is significant (b = 0.176, t-value = 1.771, p-value = 0.039). Figure 3 shows the moderation relationship. The results show that the nurses with an external locus of control are more likely to engage in CWB when faced with injustice compared to those with an internal locus of control.

Discussion

This study serves as an extension to the literature concerning CWB by not only proposing, but also testing, a comprehensive model of the contingencies that influence the

© 2019 Australian HR Institute 19 Asia Pacific Journal of Human Resources perceptions among nurses about injustice when subjected to abusive supervision, as well as the contingencies that affect their decision to engage in CWB after perceiving injustice. Furthermore, the mediating role of nurses’ perception of injustice has been examined between abusive supervision and CWB, as well as moderators, power distance orientation and locus of control.

5

4.5

4

3.5 Low PDO 3 NPI High PDO 2.5

2

1.5

1 Low AS High AS

Figure 2 Moderation effect of power distance orientation on the relationship between abusive supervision and nurses’ perception of injustice

5

4.5

4

3.5 Low LC 3

CWB High LC 2.5

2

1.5

1 Low NPI High NPI

Figure 3 Moderation effect of locus of control on the relationship between nurses’ perception of injustice and counterproductive work behaviors (CWB)

20 © 2019 Australian HR Institute Yoke May Low et al.

This empirical study has provided supporting evidence linking abusive supervision and CWB among nurses in public hospitals. Although Tepper (2007) has mentioned that the healthcare industry appears to be particularly prone to abusive behaviors, the impact of abusive supervision on nurses’ perception of injustice and CWB has yet to be acknowledged and studied extensively within this group of nurses. This study has filled this gap. The findings of this study show that abusive supervision is an important determinant of CWB. Based on previous studies, abusive supervision is related to intention to quit (Schat, Frone and Kelloway 2006; Tepper 2000), deviance (Schaubhut, Adams and Jex 2004; Tepper 2007), supervisor directed aggression (Dupre et al. 2006; Inness, Barling and Turner 2005), as well as retaliatory behaviors (Mitchell and Ambrose 2007; Skarlicki, Folger and Tesluk 1999). This supervisory mistreatment has been found to promote vari- ous negative behaviors among subordinates, giving them leeway to seek redress for their poor treatment. Moreover, as outlined by social exchange theory, the exchanges that take place are largely seen as interdependent and contingent on rewarding reactions from others in the social relationship, and thus generate obligations (Blau 1964). From a differ- ent perspective, CWB reflects social behavior as a result of a process that involves verbal transactions or exchanges (i.e. abusive supervision). This dictates that the process mani- fested in social exchange theory furthers our understanding of the reasons for an employee to engage in CWB solely due to abusive supervision. Hence, one can say that the present study has contributed to the body of knowledge by adding evidence of the predictive validity of abusive supervision on CWB. The descriptive statistics show that abusive super- vision is experienced by nurses in public hospitals, and consequently these abused nurses are likely to engage in CWB. The results obtained from the current study are consistent with those of previous studies, that abusive supervision leads to perceived injustice among abused individuals (Aryee et al. 2007; Tepper 2000). As well, it has been found that the abusive supervision experienced by nurses in public hospitals is admittedly an unjust act that influences their perception of justice towards their supervisors. Hence, it is evident that the degree to which supervisors engage in abusive behavior does affect the perceptions the nurses have towards injustice. Moreover, supervisors who frequently act abusively towards their nurses instill a certain feeling of discrimination in them. Such a feeling will in turn influence their decisions to respond and react to such acts of abusive supervision. Thus, when a sense of discrimination surfaces, these abused nurses may have a strong tendency to engage in a counterproductive act in order to restore justice. In general, the result of the current study is in line with those of prior scholars who have asserted that perception of injustice mediates the relationship between abusive supervision and various organizational outcomes (Bies and Tripp 2001; Eisenberger et al. 2004; Greenberg and Alge 1998; Skarlicki, Folger and Tesluk 1999; Tepper 2000). This further justifies the result of the current study regarding the engagement in CWB of abused nurses in public hospitals. The result also portrays that abused nurses only engage in CWB when they perceive injustice. The perceived injustices resulting from abusive supervision therefore prompt an abused nurse to seek a restoration of justice. As a result, nurses will reciprocate

© 2019 Australian HR Institute 21 Asia Pacific Journal of Human Resources by engaging in CWB in order to even the score. This is also consistent with Adam’s (1963) equity theory, which explains that the sense of fairness is dependent on the comparison a subordinate makes between his or her reward with that received by others in a similar situation. Hence, abused nurses begin comparing themselves with their peers when they are abused by their supervisors in order to make justice judgments. When a sense of discrimination arises, these abused nurses might have a strong tendency to engage in counterproductive acts in order to restore justice. This implies that nurses’ perception of injustice plays a mediating role in the relationship between abusive supervision and CWB among nurses from public hospitals. Our study has shown that if the power distance orientation is high, abusive super- vision may not translate to the perception of injustice among nurses. The characteristics of the nurses who participated in this study indicate that nurses employed in public hospi- tals have a ‘moderate’ level of power distance orientation (mean score = 2.56). This sug- gests that abusive supervision may translate to a moderate level of perception of injustice among nurses. Thus, this study has successfully shown that power distance orientation has a moderating effect upon the relationship between abusive supervision and perception generated by nurses from public hospitals towards injustice, which parallels the findings reported by Wang et al. (2012). The result shows that the relationship between abusive supervision and nurses’ perception of injustice can be further exacerbated by power dis- tance orientation among nurses from public hospitals. Nurses with low power distance orientation have a greater tendency to react negatively to abusive supervision, as they are less willing to accept the unequal distribution of power rendered by their supervisors. In contrast, nurses with higher power distance orientation are less likely to perceive the abu- sive act as unjust, because they are more dependent on their leaders for resources and guidance (Wang et al. 2012). Our study has also shown that nurses with an internal locus of control are less likely to engage in CWB, even if they perceive injustice in the way they are being abused. The nurses employed in public hospitals have an internal locus of control (mean score = 2.52). The cumulative effect of these moderators indicates that nurses in public hospitals do engage in CWB to ‘some’ extent. In general, the results obtained from the current study are consistent with the findings retrieved from prior studies, which showed that individuals with an external locus of control have more tendencies to increase the likelihood of negative behavior (Spector and Fox 2002; Wei and Si 2013). Hence, one can rationalize that nurses in public hospitals with an external locus of control are likely to respond to injustice when subjected to abusive supervision by engaging in CWB, as they have no control over the unpleasant incidents and have no ability to influence discriminatory acts. Therefore, nurses with an external locus of control, especially those who are disgruntled (i.e. perceive injustice) in their work environment (i.e. experiencing abusive supervision), are more likely to act counterpro- ductively. The descriptive statistics on CWB indicate that the nurses have engaged in CWB a few times a year. Our study has demonstrated that the impact of abusive supervision on CWB can be clearly articulated if the moderators (power distance

22 © 2019 Australian HR Institute Yoke May Low et al. orientation and locus of control) and mediator (nurses’ perception of injustice) are accounted for. The findings obtained from this study provide insights into nurse managers and organizations as to how abusive supervision can affect a nurse’s decision to engage in CWB. Additionally, since abusive supervision and CWB are prevalent in public hospitals, motivational or stress seminars can be provided to all staff to help them cope with stress. The occurrences of abusive behavior are usually due to a stressful work environment, which has been proven to be harmful to both the organization and its members. Furthermore, the findings generated from this study signify that abused nurses who perceive abusive supervision as mistreatment are more motivated to engage in CWB. The hospital management thus needs to take the necessary actions to eliminate all forms of abusive supervisory behaviors. Training should be provided to immediate nurse supervisors in order to create an awareness of the harmful effects of abusive supervision on nurses. It is vital to eliminate all forms of abusive supervision, as nurses may fail to cope with the excessive stress in the workplace caused by high workload and mistreatment. The evidence of the significant moderators for the relationships between abusive supervision, nurses’ perception of injustice and CWB suggests that supervisory-focused training tailored to the needs of individual nurses can help supervisors understand why nurses engage in CWB, in a way that accommodates the nurses colorful palette of values and personalities. For instance, nurse supervisors should manage their subordinates based on individual cultural orientation, as people present with a varied set of individual cultural values as opposed to the culture of a certain nation. Spector (2011) also has raised awareness of the potentially useful association between personality and CWB by supervisors, as leadership practices can be customized based on the needs of individual employees via supervisory-focused training to understand why employees engage in CWB. Finally, the results also suggest that hospitals should consider the roles of power distance orientation and locus of control in selecting nurses for employment. Hiring nurses with high power distance orientation and internal locus of control can help diminish the perception of injustice and occurrences of CWB due to mistreatment/abusive supervision, to which those in the healthcare industry are especially vulnerable (Tepper 2007). Public hospital administrators should execute training programs for nurses to overcome both abusive treatment and environments, as well as providing a grievance system where they can express themselves and so mitigate any CWB.

Conclusions, limitations, and directions for future research

This study reveals the significant associations between abusive supervision, nurses’ percep- tion of injustice, power distance orientation, locus of control and CWB. The results of this study (among nurses in public hospitals) indicate that nurses who are subjected to abusive supervision are more likely to engage in CWB. The nurses perceive injustice when sub- jected to abusive supervision and this perception leads to their behavioral outcomes. This study has shown that nurses with a high level of power distance orientation are less likely to perceive abusive supervision as injustice, while those with an internal locus of control are less likely to engage in CWB.

© 2019 Australian HR Institute 23 Asia Pacific Journal of Human Resources

This study is not without limitations. First, the data for this study were collected from a single source at a single point in time. Furthermore, since the data were collected from a single source, this could cause a common variance bias (Podsakoff et al. 2003). Second, the cross-sectional research design adopted in this study did not permit causal relation- ships to be discussed in detail. Longitudinal data gathered from multiple sources should be able to address this limitation. Third, the sample respondents (nurses) were from pub- lic hospitals. The situation in private hospitals could be different. Future research can con- sider 1) longitudinal study, 2) inclusion of nurses from private hospitals, and 3) collection of data from multiple sources.

Yoke May Low completed her PhD at Universiti Putra Malaysia in 2017. Her area of interest is Human Resources and OB.

Murali Sambasivan is a Professor of Management Science at Faculty of Business and Law at Taylor’s University, Lakeside Campus, Malaysia and an Adjunct Professor at Thiagarajar School of Management, Madurai, India. He has a PhD from University of Alabama, USA. He has published in many international journals of repute.

Jo Ann Ho is an Associate Professor at Universiti Putra Malaysia. She has a PhD from Cardiff Busi- ness School. Her areas of interest are CSR, HR and OB.

References

Adams JS (1965) Inequity in social exchange. In: Berkowitz L (ed) Advances in experimental social psychology, Vol. 2, 267–299. Academic Press, New York. Ambrose ML, MA Seabright and M Schminke (2002) Sabotage in the workplace: the role of organi- zational injustice. Organizational Behavior and Human Decision Processes 89(1), 947–965. Aryee S, ZX Chen, LY Sun and YA Debrah (2007) Antecedents and outcomes of abusive supervi- sion: test of a trickle-down model. Journal of Applied Psychology 92(1), 191–201. Ashforth BE (1997) Petty tyranny in organizations: a preliminary examination of antecedents and consequences. Canadian Journal of Administrative Sciences 14(2), 126–140. Becker J, K Klein and M Wetzels (2012) Hierarchical latent variable models in PLS-SEM: guidelines for using reflective-formative type models. Long Range Planning 45(5–6), 359–394. Begley TM, C Lee, Y Fang and J Li (2002) Power distance as a moderator of the relationship between justice and employee outcomes in a sample of Chinese employees. Journal of Manage- rial Psychology 17(8), 692–711. Bennett RJ and SL Robinson (2000) Development of a measure of workplace deviance. Journal of Applied Psychology 85(3), 349–360. Berry CM, DS Ones and PR Sackett (2007) Interpersonal deviance, organizational deviance and their common correlates: a review and meta-analysis. Journal of Applied Psychology 92(2), 410–424. Bies RJ (1999) Interactional (in)justice: the sacred and the profane. In Greenberg J and Cropanzano R (eds) Advances in organizational behavior. New Lexington Press, San Francisco, CA. Bies RJ and TM Tripp (1996) Beyond distrust: ‘Getting even’ and the need for revenge. In Kramer RM and Tyler T (eds) Trust in organizations. Sage Publications, Newbury Park, CA.

24 © 2019 Australian HR Institute Yoke May Low et al.

Bies RJ and TM Tripp (1998) Revenge in organizations: the good, the bad, and the ugly. In Griffin RW, O’Leary-Kelly A and Collins JM (eds) Dysfunctional behavior in organizations: violent and deviant behavior. JAI Press, Stamford, CT. Bies RJ and TM Tripp (2001) A passion for justice: the rationality and morality of revenge. In Cropanzano R (ed) Justice in the workplace, Vol. II. Lawrence Erlbaum Associates, Mahwah, NJ. Blau P (1964) Exchange and power in social life. Wiley, New York, NY. Bolin A and L Heatherly (2001) Predictors of employee deviance: the relationship between bad atti- tudes and bad behavior. Journal of Business and Psychology 15(3), 405–418. Bollen KA (2011) Evaluating effect, composite and causal indicators in structural equation models. MIS Quarterly 35(2), 359–372. Bowling NA and TA Beehr (2006) Workplace harassment from the victim’s perspective: a theoreti- cal model and meta-analysis. Journal of Applied Psychology 91(5), 998–1012. Bryman A and E Bell (2007) Business research methods, 2nd edn. Oxford University Press, New York, NY. Burton JP and JM Hoobler (2011) Aggressive reactions to abusive supervision: the role of interac- tional justice and narcissism. Scandinavian Journal of Psychology 52(4), 389–398. Chappell D and V Di Martino (2006) Violence at work, 3rd edn. International Labour Office, Geneva. Clugston M, JP Howell and PW Dorfman (2000) Does cultural socialization predict multiple bases and foci of commitment? Journal of Management 26(1), 5–30. Cohen J (1988) Statistical power analysis for the behavioral sciences, 2nd edn. Lawrence Earlbaum Associates, Hillsdale, NJ. Colquitt JA (2001) On the dimensionality of organizational justice: a construct validation of a mea- sure. Journal of Applied Psychology 86(3), 386–400. Colquitt J, J Greenberg and C Zapata-Phelan (2005) What is organizational justice?: a historical overview. In Greenberg J and Colquitt JA (eds) The handbook of organizational justice,3–56. Lawrence Erlbaum Associates, Mahwah, NJ. Crooker KJ, FL Smith and F Tabak (2002) Creating work-life balance: a model of pluralism across life domains. Human Resource Development Review 1(4), 387–419. Davis WL and DE Davis (1972) Internal and external control and attribution of responsibility for success and failure. Journal of Personality 40, 123–126. Detert JR, LK Trevino, ER Burris and M Andiappan (2007) Managerial modes of influence and counterproductivity in organizations: a longitudinal business-unit-level investigation. Journal of Applied Psychology 92(4), 993–1005. Dorfman PW and JP Howell (1988) Dimensions of national culture and effective leadership pat- terns: Hofstede revisited. Advances in International Comparative Management 3(1), 127–150. Dupre KE, M Inness, CE Connelly, J Barling and C Hoption (2006) Workplace aggression in teen- age part-time employees. Journal of Applied Psychology 91(5), 987–997. Edward KL, W Ousey, P Warelow and S Lui (2014) Nursing and aggression in the workplace: a sys- tematic review. British Journal of Nursing 23(12), 653–659. Eisenberger R, P Lynch, J Aselage and S Rohdieck (2004) Who takes the most revenge? Individual differences in negative reciprocity norm endorsement. Personality and Social Psychology Bulletin 30(6), 787–799. Farh JL, RD Hackett and J Liang (2007) Individual-level cultural values as moderators of per- ceived organizational support-employee outcome relationships in China: comparing the

© 2019 Australian HR Institute 25 Asia Pacific Journal of Human Resources

effects of power distance and traditionality. The Academy of Management Journal 50(3), 715–729. Farrell GA, C Bobrowski and P Bobrowski (2006) Scoping workplace aggression in nursing: findings from an Australian study. Journal of Advanced Nursing 55(6), 778–787. Fox S and PE Spector (1999) A model of work frustration-aggression. Journal of Organizational Behavior 20(6), 915–931. Giacalone RA and JE Greenberg (1997) Antisocial Behavior in Organizations. Sage Publications Inc, Thousand Oaks, California. Greenberg J and BJ Alge (1998) Aggressive reactions to workplace injustice. In Griffin RW, O’Leary-Kelly A and Collins JM (eds) Dysfunctional behavior in organizations: violent and devi- ant behavior. JAI Press, Stamford, CT. Greenberg J and KS Scott (1996) Why do workers bite the hands that feed them? Employee theft as a social exchange process. Research in Organizational Behavior 18, 111–156. Griffin RW and AM O’Leary-Kelly (2004) The dark side of organizational behavior. John Wiley & Sons, San Francisco, CA. Hair JF, GTM Hult, CM Ringle and M Sarstedt (2014) A primer on partial least squares structural equation modeling (PLS-SEM). Sage, Thousand Oaks, CA. Henseler J, CM Ringle and M Sarstedt (2015) A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of Academic Marketing Science 43, 115–135. Hershcovis MS and J Barling (2007) Towards a relational model of workplace aggression. In: Lan- gan-Fox J, Cooper CL, Klimoski RJ (eds) Research Companion to the Dysfunctional Workplace: Management Challenges and Symptoms,68–284. Edward Elgar, Northampton, MA. Hockley C (2002) Silent hell: workplace violence and bullying. Peacock Publications, Norwood, OH. Hofstede G (1980) Culture and organizations. International Journal of Management & Organization 10(4), 15–41. Hofstede G (2001) Culture’s consequences: comparing values, behaviors, institutions and organizations across nations. Sage, Thousand Oaks, CA. Hofstede G and MH Bond (1984) Hofstede’s culture dimensions: an independent validation using Rokeach’s value survey. Journal of Cross-Cultural Psychology 15(4), 417–433. Hogan J and R Hogan (1989) How to measure employee reliability. Journal of Applied Psychology 74, 273–279. Hulland J (1999) Use of Partial Least Squares (PLS) in Strategic Management Research: A Review of Four Recent Studies. Strategic Management Journal 20, 195–224. Hutchinson M, M Vickers, D Jackson and L Wilkes (2006) Workplace bullying in nursing: towards a more critical organizational perspective. Nursing Inquiry 13(2), 118–125. Inness M, J Barling and N Turner (2005) Understanding supervisor-targeted aggression: a within- person, between-jobs design. Journal of Applied Psychology 90(4), 731–739. Ivancevich JM, TN Duening, JA Gilbert and R Konopaske (2003) Deterring white-collar crime. The Academy of Management Executive 17(2), 114–127. Jarvis CB, SB MacKenzie and PM Podsakoff (2003) A critical review of construct indicators and measurement model misspecification in marketing and consumer research. Journal of Consumer Research 30(2), 199–218. de Jonge J and MCW Peeters (2009) Convergence of self-reports and co-worker reports of CWB: a cross sectional multi-source survey among healthcare workers. International Journal of Nursing Studies 46(5), 699–707.

26 © 2019 Australian HR Institute Yoke May Low et al.

Kirkman BL and DL Shapiro (2001) The Impact of cultural values on job satisfaction and organiza- tional commitment in self managing work teams: The mediating role of employee resistance. The Academy of Management Journal 44(3), 557–569. Knorz C and D Zapf (1996) Mobbing: an extreme form of social stressors at work. Zeits- chriftfurArbeits-€ und Organisationspsychologie 40, 12–21. Krejcie RV and DW Morgan (1970) Determining sample size for research activities. Educational and Psychological Measurement 30, 607–610. Lam SSK, J Schaubroeck and S Aryee (2002) Relationship between organizational justice and employee work outcomes: a cross-national study. Journal of Organizational Behavior 23(1), 1–18. Lee C, M Pillutla and KS Law (2000) Power distance, gender and organizational justice. Journal of Management 26(4), 685–704. LePine JA, A Erez and DE Johnson (2002) The nature and dimensionality of organizational citizen- ship behavior: a critical review and meta-analysis. Journal of Applied Psychology 87(1), 52–65. Levine EL (2010) Emotion and power (as social influence): their impact on organizational citizen- ship and counterproductive individual and organizational behavior. Human Resource Manage- ment Review 20(1), 4–17. Marcus B and H Schuler (2004) Antecedents of counterproductive behavior at work: a general per- spective. Journal of Applied Psychology 89(4), 647–660. Martin J (1981) Relative deprivation: a theory of distributive injustice for an era of shrinking resources. In Cummings LL and Staws BM (eds) Research in organizational behavior,53–107. JAI Press, Greenwich, CT. Martinko MJ, MJ Gundlach and SC Douglas (2002) Toward an integrative theory of counterpro- ductive workplace behavior: a causal reasoning perspective. International Journal of Selection and Assessment 10(1/2), 36–50. Martinko MJ, P Harvey, JR Brees and J Mackey (2013) A review of abusive supervision research. Journal of Organizational Behavior 34, 120–137. Mayhew C and D Chappell (2001) Occupational violence types, reporting patterns and variations between health sectors. Taskforce on the Prevention and Management of Violence in the Work- place, Discussion Paper No. 1. University of NSW, Kensington. Mitchell MS and ML Ambrose (2007) Abusive supervision and workplace deviance and the moder- ating effects of negative reciprocity beliefs. Journal of Applied Psychology 92(4), 1159–1168. Mitchell MS and ML Ambrose (2012) Employees’ behavioural reactions to supervisor aggression: an examination of individual and situational factors. Journal of Applied Psychology 97(6), 1148– 1170. Mount M, R Ilies and E Johnson (2006) Relationship of personality traits and counterproductive work behaviors: the mediating effects of job satisfaction. Personnel Psychology 59(3), 591–622. Murray JS (2009) Workplace bullying in nursing: a problem that can’t be ignored. Medsurg Nursing Sept/Oct, 273–276. Nicholson JD and LP Stepina (1998) Cultural values: a cross-national study. Cross Cultural Manage- ment: An International Journal 5(1/2), 33–47. O’Leary-Kelly AM, RW Griffin and DJ Glew (1996) Organization-motivated aggression: a research framework. Academy of Management Review 21(1), 225–253. Penney LM and PE Spector (2002) Narcissism and counterproductive work behavior: do big- ger egos mean bigger problems? International Journal of Selection and Assessment 10(1/2), 126–134.

© 2019 Australian HR Institute 27 Asia Pacific Journal of Human Resources

Phares EJ, KG Wilson and NW Klyver (1971) Internal-external control and the attribution of blame under neutral and distractive conditions. Journal of Personality and Social Psychology 18(3), 285–288. https://doi.org/10.1037/h0031007. Pillay S (2017) Malaysia’s medical tourism set to soar, hits RM1bil mark last year. News Strait Times, Kuala Lumpur, Malaysia Retrieved from http://www.nst.com.my/news/2017/04/226430/ malaysias-medical-tourism-set-soar-hits-rm1bil-mark-lastyear. Podsakoff PM, SB MacKenzie, JY Lee and NP Podsakoff (2003) Common method biases in behav- ioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology 88(5), 879–903. Podsakoff NP, SW Whiting, PM Podsakoff and BD Blume (2009) Individual-and organizational- level consequences of organizational citizenship behaviors: a meta-analysis. Journal of Applied Psychology 94(1), 120–143. Preacher KJ and AF Hayes (2008) Asymptotic and resampling strategies for assessing and compar- ing indirect effects in multiple mediator models. Behavior Research Methods 40(3), 879–891. Quick JC and DL Nelson (2009) Principles in organizational behavior: realities and challenges (Inter- national Edition), 6th edn. South-Western, Nashville, TN. Raman P, M Sambasivan and N Kumar (2016) Counterproductive work behavior among front- line government employees: role of personality, emotional intelligence, affectivity, emotional labor, and emotional exhaustion. Journal of Work and Organizational Psychology 32(1), 25– 37. Robinson SL and J Greenberg (1998) Employees behaving badly: dimensions, determinants and dilemmas in the study of workplace deviance. In Cooper CL and Rousseau DM (eds) Trends in organizational behavior, Vol. 5, 1–30. New York, NY, John Wiley & Sons Ltd. Rosenstein AH and M O’Daniel (2005) Original research: disruptive behavior and clinical out- comes: perceptions of nurses and physicians: nurses, physicians, and administrators say that clinicians’ disruptive behavior has negative effects on clinical outcomes. AJN The American Jour- nal of Nursing 105(1), 54–64. Rotter JB (1966) Generalized expectancies for internal versus external control of reinforcement. Psy- chological Monographs: General and Applied. 80, 1–28. https://doi.org/10.1037/h009297. Schat ACH, MR Frone and EK Kelloway (2006) Prevalence of workplace aggression in the US Workforce: findings from a National Study. In Kelloway EK, Barling J and Hurrell JJ (eds) Handbook of workplace violence,47–89. Sage, Thousand Oaks, CA. Schaubhut N, G Adams and S Jex (2004) Self-esteem as a moderator of the relationships between abu- sive supervision and two forms of workplace deviance. Society for Industrial Organizational Psy- chology, Chicago, IL. Sekaran U (2000) Research Methods for business: A Skill Building Approach. John Wiley & Sons Inc, Singapore. Skarlicki DP and R Folger (1997) Retaliation in the workplace: the roles of distributive, procedural, and interactional justice. Journal of Applied Psychology 82(3), 434–443. Skarlicki DP, R Folger and P Tesluk (1999) Personality as a moderator in the relationship between fairness and retaliation. Academy of Management Journal 42(1), 100–108. Spector PE (1988) Development of the work locus of control scale. Journal of Occupational Psychol- ogy 61(4), 335–340. Spector PE (2011) The relationship of personality to counterproductive work behavior (CWB): an integration of perspectives. Human Resource Management Review 21, 342–352.

28 © 2019 Australian HR Institute Yoke May Low et al.

Spector PE and S Fox (2002) An emotion-centered model of voluntary work behavior: some paral- lels between counterproductive work behavior and organizational citizenship behavior. Human Resource Management Review 12(2), 269–292. Spector PE and S Fox (2005) The stressor-emotion model of counterproductive work behavior. In Spector PE and Fox S (eds) Counterproductive workplace behavior: investigations of actors and targets. APA Press, Washington, DC. Spector PE and S Fox (2010) Counterproductive work behaviour and organizational citizenship behaviour. Applied Psychology 59(1), 21–39. Storms PL and PE Spector (1987) Relationships of organizational frustration with reported beha- vioural reactions: the moderating effect of locus of control. Journal of Occupational Psychology 60(3), 227–234. Tepper BJ (2000) Consequences of abusive supervision. Academy of Management Journal 43(2), 178–190. Tepper BJ (2007) Abusive supervision in work organizations: review, synthesis, and research agenda. Journal of Management 33(3), 261–289. Tepper BJ, MK Duffy, CA Henle and LS Lambert (2006) Procedural injustice, victim precipitation, and abusive supervision. Personnel Psychology 59(1), 101–123. Thau S, RJ Bennett, MS Mitchell and MB Marrs (2009) How management style moderates the rela- tionship between abusive supervision and workplace deviance: an uncertainty management the- ory perspective. Organizational Behavior and Human Decision Processes 108(1), 79–92. Wang W, J Mao, W Wu and J Liu (2012) Abusive supervision and workplace deviance: the mediat- ing role of interactional justice and the moderating role of power distance. Asia Pacific Journal of Human Resources 50(1), 43–60. Wei F and S Si (2013) Tit for tat? Abusive supervision and counterproductive work behavior: the moderating effects of locus of control and perceived mobility. Asia Pacific Journal of Manage- ment 30, 281–296. Zellars KL, BJ Tepper and MK Duffy (2002) Abusive supervision and subordinates’ organizational citizenship behavior. Journal of Applied Psychology 87(6), 1068–1076.

© 2019 Australian HR Institute 29