RESEARCH AND PRACTICE
IN HUMAN RESOURCE MANAGEMENT

Highlight, copy & paste to cite:

Martin, A., Gray, C. & Adam, A. (2007). Nurses’ Responses to Workplace Verbal Abuse: A Scenario Study of the Impact of Situational and Individual Factor, Research and Practice in Human Resource Management, 15(2), 41-61.

Nurses’ Responses to Workplace Verbal Abuse: A Scenario Study of the Impact of Situational and Individual Factor

Angela Martin, Cynthia Gray & Andrea Adam

Abstract

Verbal abuse in the workplace represents a costly problem for human resource management, with implications for staff productivity and retention. Nurses are a profession exposed to extremely high levels of verbal abuse, particularly from patients, but also from colleagues, making these abusive behaviours, and reactions to them, of vital significance to hospital personnel management professionals. The present study investigated the extent to which responses to verbal abuse scenarios varied as a function of the role and gender of the perpetrator, the situation in which the verbal abuse occurred, and the participant’s level of assertiveness, with a view to establishing key contributors for management attention. Participants were 62 female registered nurses. Results showed evidence of variation in cognitive, affective and behavioural attitudes toward the verbal abuse described, particularly as a function of situation. In particular, verbal abuse was reported to be more frequent, considered less acceptable and elicited more affective arousal in non emergency than in emergency situations. These findings were consistent with predictions that were based on attribution theory. Responses to verbal abuse also varied in relation to the nurse’s assertiveness, which is a finding suggesting that assertiveness training is a potential strategy for protecting employees against the effects of verbal abuse. This study has important implications for managers in hospitals, and indeed, other organisational settings in that dealing with verbal abuse requires cognisance of the need for the implementation of both organisational and individual levels of intervention, within a framework that acknowledges the complexity of responses to verbal abuse in the workplace.

Introduction

Workplace violence is an issue of global concern (Gill, Fisher & Bowie 2002), with potential negative ramifications for organisations including reduced productivity and morale, and increased absenteeism and turnover rates. It has been suggested that nurses are subjected to up to three times as much violence in the workplace than any other profession (Paterson, McCornish & Bradley 1999, Perrone 1999). Verbal abuse is an insidious form of workplace violence that is a major contributor to dissatisfaction and high rates of attrition in nurses (Vogt, Cox, Velthouse & Thames 1983, Bush & Gilliland 1995, Smith 1997, Sofield & Salmond 2003). Although patients are a major source of workplace violence for nurses, it is noticeable nurses have also reported workplace violence from colleagues, with incidences of up to 36 per cent reported in private hospitals (Hegney, Plank & Parker 2003). This high rate of workplace violence towards nurses makes verbal abuse a significant concern for human resource management within hospital settings.

In order to understand the specifics of the nurses’ experiences of workplace verbal abuse, conceptualisations need to incorporate analysis of the hospital context. It has been hypothesised that the hierarchical nature of the hospital environment may contribute to high rates of verbal abuse through its hierarchical organisation and exposure to suffering and grief (Simms 2000). In addition, it is possible that workplace violence towards nurses is exacerbated by the traditional gender inequalities across roles of differing status that exist within hospital settings, where the lower status role of nurses has been compounded with a gender that has also traditionally accorded lower status. However, there is limited understanding of how gender interacts with the status differences between doctors and nurses to influence the frequency of, and reactions to, verbal abuse. In turn, this lack of information constrains understanding of the way changing gender distributions within nursing and doctoring professions might affect the experienced verbal abuse of nurses.

Empirical research into workplace violence, which might aid the development of strategies to deal effectively with abuse, has had practical difficulties (Douglas & Martinko 2001). This condition has resulted in the majority of literature being atheoretical and anecdotal in nature (Cusack 2000). In contrast, the current study was underpinned by several theoretical approaches that provide insight into the verbal abuse behaviour of doctors and nurses, including: Social Role, Social Identity, and Attribution theories. Furthermore, while the prevalence of verbal abuse may be relatively well known, the scenario methodology, that was adapted for this investigation, provides a format to examine workplace violence, in this case verbal abuse, in an empirical and replicable manner. This method also allowed systematic rather than incidental investigation into the impact that situational variables have on incidents of verbal abuse.

The aim of this study was to investigate how nurses’ reactions to verbal abuse are influenced by the role and gender of the abuser, and the situation in which abuse occurs. Because assertiveness training is a popular strategy currently promoted to deal with workplace violence (e.g., Milstead 1996, Mimura & Griffiths 2003), the ways in which the nurses’ assertiveness affects their experience of, and their likely reactions to, verbal abuse, was also examined. Several theoretically based predictions about nurses’ likely responses to verbal abuse enacted in a range of scenarios are developed in the following literature review. The remainder of the paper is structured in a scientific report format, culminating in a discussion of the findings in relation to the theoretical predictions that were made, and the implications for human resource management.

Literature Review

The Problem of Verbal Abuse

Verbal abuse in organisations is a central feature of the conceptualisations of counter productive work behaviour, workplace incivility, aggressive interpersonal behaviour, overt bullying and workplace violence (Arway 2002, Barron 2002, Fox & Spector 2005). Verbal abusers use aggressive communication tactics such as humiliation, sarcasm, insults, labelling, and blaming in an attempt to discredit the victim. It has been argued that verbal abuse is a less extreme, but more widespread form of workplace violence that has been empirically neglected in comparison with the study of physical attacks (Gill, et al. 2002).

The occurrence of verbal abuse has serious implications not only for the recipients of such communication, but also for the organisations in which the abuse occurs. An incident of verbal abuse can potentially reduce productivity, efficiency, and morale and could lead to increased absenteeism, higher turnover rates, greater recruitment costs, elevated workers’ compensation premiums, and possible legal defence costs (Queensland Government 2001). Affective responses to verbal abuse can vary from neutral, or unperturbed, to very distressed, potentially resulting in severe emotional impairment and reduced self esteem (Elgin 1980). Durkin and Wilson (1999) argued that verbal abuse could be as distressing as a physical attack. Antai-Otong (2001) found that the accumulated stress related to repeated experiences of verbal aggression was not only associated with psychological complaints such as depression, panic disorder, and posttraumatic stress disorder, but also could be related to physical ailments such as hypertension. Of particular importance for human resource managers, in healthcare settings, is the research has demonstrated that verbal abuse is one of the strongest factors that contributes to dissatisfaction and high rates of attrition in nurses (Vogt, et al. 1983, Bush & Gilliland 1995, Smith 1997, Sofield & Salmond 2003).

Role and Gender Influences on Responses to Verbal Abuse

Nurses have frequently described their work environment as a hostile climate where scapegoating, disrespectful treatment, and lack of support are commonplace (Smith, Droppleman & Thomas 1996). Healthcare settings are thought to be predisposed to verbal abuse behaviours by the frequent occurrence of high level stressors such as failure to cure, suffering, and death. It has been theorised that unexpressed grief in staff may surface as aggressive behaviours such as verbal abuse (Simms 2000). It has also been argued that the potential for verbal abuse rises when collaboration and communication are overridden by hierarchical structures, such as those observed within hospitals (Simms 2000). Organisational culture within hospitals is deeply entrenched in a hierarchical structure, which recognises doctors as the pinnacle of the organisation and the sole possessors of power (Numerof 1978, McCall 1996, Smith, et al. 1996). This power is derived from a number of sources such as expert and legitimate power invested by the organisation, as well as power drawn from gender inequality, reflecting the fact that historically the majority of doctors have been male and the majority of nurses have been female (Pokalo 1991, Woodward & House 1997, Worchel, Cooper, Goethals & Olsen 2000). Thus, a feminist analysis links hospital culture to the broader patriarchal structure that operates to subordinate women in society. It is argued that verbal abuse is one strategy used by men as a method of asserting and maintaining this dominance (Bruder 2001).

Another important concept in theorising aggressive behaviours such as verbal abuse is that of gender roles. This notion advances the socially determined classifications of ‘appropriate’ behaviour as prescribed by gender. According to Social Role theory, gender roles are evident in social expectations that women should be communally oriented and primarily concerned with the welfare of others, whilst men should display more ‘agentic’ characteristics such as independence and assertiveness (Eagly 1987). The social role of the nurse aligns with traditional female roles through caring for the sick, whilst functioning as a supportive labour force for doctors. This role is described by Muff (1982) as involving and following the orders of doctors, running their errands, and tolerating their abuse and rudeness.

It could be argued that the power gradient between men and women is fading. Recent social changes have reflected an increasing convergence of gender roles and an increase in women’s access to previously male dominated roles, such as medicine, and the power associated with those roles (Diekman & Eagly 2000). However, Inglehart and Norris (2003) noted that women continue to predominately hold jobs of lower status and rewards, and recent literature shows that very little has changed in the area of gender roles for doctors and nurses (Simms 2000, Bruder 2001, Cook, Green & Topp 2001, Dunn 2003, Sofield & Salmond 2003).

An analysis of verbal abuse informed by issues of power, status and social roles support the expectation that this behaviour is more likely to occur with doctors as perpetrators and nurses as victims (known as vertical violence). Substantial literature has documented that doctors are a frequent source of verbal abuse attacks on nurses (Lopez 1993, Begany 1995, Farrell 1999, Simms 2000, Bruder 2001, Cook, et al. 2001, Sofield & Salmond 2003, Buback 2004). For example, one study reported that two thirds of nurses acknowledged they had experienced verbal abuse from a doctor in the previous twelve months (Begany 1995). However, it is important to note that the verbal abuse of nurses is not restricted to doctors alone. Hegney, et al. (2003) also reported that apart from patients and doctors, other nurses are a frequent source of verbal abuse. The phenomenon where employees in similar roles become the instigators of aggression and violence towards each other was first investigated amongst nurses, and became known as horizontal violence (Taylor 2001, Bowie 2002). Horizontal violence between nurses has long been acknowledged as a negative adaptation of oppressed group behaviour related to the subordination of women within the health care system (Duffy 1995, McCall 1996, Smith, et al. 1996, Roberts 1997, Dunn 2003). Freire’s (1970) model of oppression argued that subordinate groups learn to value the norms of the dominant group whilst simultaneously learning to abhor their own attributes. As evidence of this, when nursing leaders emerge, they inevitably adopt the values of the dominant group (doctors) which results in internal domination of nurses by members of their own group (Roberts 1983, Duffy 1995). Further, Cox (1994) suggested that verbal abuse by nurses that is directed at other nurses and perceived subordinates, could be seen as a coping mechanism related to frustration and negative self esteem as a consequence of being treated as inferior members of the healthcare team.

Another theoretical perspective relevant to exploring the reactions to verbal abuse amongst doctors and nurses is Social Identity theory, which has made an important contribution to understanding the phenomena of social group behaviour (Tajfel & Turner 1979). In this framework it is argued that individuals see themselves as members of groups and they self categorise as a means of constructing and maintaining identity. Perceived membership of a group, the ingroup, is defined in relation to an outgroup. Ingroup members tend to be characterised with positive qualities and outgroup members are characterised with negative qualities (Worchel, et al. 2000). The purpose of this discrimination is to maintain or realise supremacy over an outgroup (Tajfel & Turner 2001). Social Identity Theory may offer an explanation of reactions to vertical and horizontal verbal abuse. It could be suggested that ingroup members (nurses or females — horizontal perpetrators) tend to perceive that outgroup members (doctors or males — vertical perpetrators) have negative qualities. Thus, verbal abuse from an outgroup member may fit the concept of negative qualities of the outgroup, and although offensive, it would not be an unexpected outcome. On the other hand, verbal abuse instigated by a member of one’s own ingroup (other nurses, or other women) would be considered more offensive, as ingroup members have higher expectations of positive characteristics amongst members and may include beliefs that members should respect one another and demonstrate ‘solidarity’. In accord with this theoretical rationale, and in particular Social Identity theory, several predictions were made in relation to female nurses’ responses to verbal abuse as a function of the gender of the actor. Although traditionally there is some overlap between gender and status (operationalised here as role), an increasing number of women are entering the medical profession changing the distribution of genders. Therefore, role and gender were treated as separate variables. From this theoretical underpinning the hypotheses H1 and H2 were conjectured.

H1: If the actor of verbal abuse was also a female (i.e., the same gender), it would be considered less acceptable and more upsetting, but the participant would be more likely to report an intention to deal with it assertively as status issues are less relevant.

H2: If the actor of verbal abuse was also a nurse (i.e., the same role) it would be considered less acceptable and more upsetting than if the actor was a doctor, but the participant would be more likely to report an intention to deal with it assertively as status issues are less relevant.

Other Situational Influences on Responses to Verbal Abuse

Another important contribution to developing an understanding of responses to verbal abuse is offered by Attribution theory (Heider 1958), which is increasingly utilised to explain behaviour in the workplace (Ashkanasy & Gallois 1994, Douglas & Martinko 2001). A pivotal notion of this theory is that people ascribe a particular behaviour a person exhibits as either a function of personal disposition (an internal attribution) or induced by the situation in which the behaviour occurs (external attribution). Research has shown that in conflict situations individuals demonstrate a strong propensity to analyse the cause of other people’s behaviour (Fiske & Taylor 1984). If the recipient perceives that there are mitigating circumstances for an attack, an external attribution will be made, the actor is ascribed less malevolent intent (Fiske & Taylor 1984), and lower levels of arousal, such as anger, or retaliation are observed (Baron 1985). Thus, it would be reasonable to anticipate that common situational variables, such as emergencies within the hospital environment, may induce an external attribution to be assigned to the actor of verbal abuse. In this case, the perceived acceptability of the behaviour, level of emotional reaction and likely behavioural response to the behaviour would differ from an internally attributed act of verbal abuse, a more likely attribution in a non emergency situation.

Attribution theory provides theoretical underpinning for the hypothesis H3 which combines several predictions regarding respondents’ attributions about the verbal abuse scenarios. Emergency situations could be expected to induce an external attribution to be assigned to the actor of verbal abuse and influence the perceived acceptability of the behaviour, level of emotional reaction and likely behavioural response.

H3: In a non emergency situation verbal abuse would be considered less acceptable, affective arousal to verbal abuse would be greater and the intended behavioural response would be more assertive than in an emergency situation.

Individual Influences on Responses to Verbal Abuse - Assertiveness

Regardless of the situation and attributions made, verbal abuse can have a negative impact on employees and organisations. Many organisations use education and individual empowerment strategies to manage the problem. Assertiveness training programmes for nurses are seen as an effective means of enabling them to cope with verbal abuse (Numerof 1978, McIntyre, Jeffrey & McIntyre 1984). For example, Cox (1991: 33) found “…the higher the [nurse’s] rating of assertiveness in the work setting, the more likely the nurse was to rate her handling of verbal abuse as good.”. However, it should be noted that many researchers have rejected general assertiveness training on the grounds that it often fails to deal with issues of culture, status and complex social rules (Alberti & Emmons 1982, Rakos 1997, Wilson, Lizzio & Zauner 2001). In addition, although it has been considered that assertiveness training may augment an individual’s motivation, satisfaction, and self esteem (Spreitzer 1997), the effectiveness of these strategies within the organisational context has been questioned. It has been argued that evidence for the efficacy of developing individual coping strategies within organisational settings is weak (Burke 1993) and only serves to heighten the employee’s tolerance of unacceptable behaviour (De Frank & Cooper 1987). It is often suggested that the most effective means of addressing the problem of verbal abuse within the workplace is to target the organisational culture (Burke 1993).

Despite these claims, assertiveness training for nurses remains a central part of stress management and conflict resolution as it enables them to express themselves more effectively during challenging situations and to use coping strategies (i.e., see the abuse as an indicator of a problem with the communication of the other party and not ‘take it on’ or let it be of detriment to their self esteem). Although concerns that assertiveness training fails to prevent verbal abuse from occurring have been raised, it is included in the present study due to its ubiquitous nature as a strategy for dealing with the problem of verbal abuse. Further, individuals vary in the extent to which they possess assertiveness skills and accordingly, their responses to verbal abuse could be expected to vary. Hence, it is an important individual difference variable in attempting to explain responses to the scenarios in the present research. The assertiveness literature provided the foundation for the hypotheses H4, H5 and H6 which predicted that participants’ cognitive, affective and behavioural responses to the verbal abuse will vary as a function of their level of assertiveness.

H4 Assertive participants would consider verbal abuse less acceptable and experience lower affective arousal in response.

H5: Assertive participants would be more likely to report the intention to respond actively than non assertive participants.

H6: Non assertive participants would be more likely to report the intention to respond passively than assertive participants.

These testable propositions were evaluated by simulating verbal abuse in the form of scenarios and gauging the participant’s likely responses to such events. Gender, role, and situational variables were manipulated in the scenarios and the effect of individual differences in assertiveness among participants on responses to verbal abuse was also examined. In terms of descriptive empirical evidence, the study also assessed the prevalence of verbal abuse episodes levelled at nurses (participants) by doctors (vertical violence) and by other nurses (horizontal violence) in the sample that was studied. The range of responses to the scenarios included how frequently participants had experienced behaviour similar to that described in the scenario, how acceptable they thought the behaviour was, what sort of emotional response they would be likely to have, and their intended behavioural strategy for dealing with the behaviour.

Method

Participants and Site

The participants for this study were a convenience sample of volunteers recruited from registered nurses working at an Australian private hospital. At this research site all the potential participants were all female nurses. As nursing remains a predominantly female profession (Smith, et al. 1996) it is argued that the all female sample was not problematic in terms of generalisability. Two hundred questionnaires were distributed to registered nurses via Nurse Unit Managers in six units of an Australian private hospital including the women’s and children’s unit, the surgical unit, the medical and orthopaedic unit, the operating theatre, the cardio thoracic unit, and the intensive care unit. A total of 70 questionnaires were returned, and this yielded a response rate of 32.5 per cent. The sample of female registered nurses were aged 20 to 60 years, (M = 38.87, SD = 10.67), with years of experience as a registered nurse being 0.5 to 43 years (M = 16.91, SD = 11.01), but eight participants were excluded from the study due to incomplete data on the Rathus Assertiveness Schedule, which reduced the sample size to 62.

Procedure

The completion time for the questionnaire was approximately 20 to 30 minutes. It was emphasised to the potential participants that participation was entirely voluntary and that they could withdraw from the project at any time. Confidentiality was assured as neither the questionnaires nor the envelopes contained any identifiers and each questionnaire was provided with a self sealing envelope. The participants were asked to place their completed questionnaires into the boxes provided in the staff room of each of the six units.

Measures

The pen and paper questionnaire distributed to the participants included an information sheet and a set of demographic questions such as age, and length of time working as a nurse. Also included were a series of eight scenarios, with ten questions relating to each scenario and the 30 item Rathus Assertiveness Schedule (Rathus 1973). The eight verbal abuse scenarios were adapted from vignettes used by Buback (2004), which depicted verbal abuse within a hospital operating theatre. The scenarios were confirmed as realistic and believable by a small group of experienced registered nurses who worked at the hospital that was sampled, and these nurses had skills in several specialities. These nurses were asked not to discuss the scenarios with other staff members and were excluded from the study.

Participants assessed each of the eight scenarios using ten, seven point scales. The first item assessed the frequency of experience or witness of the communication depicted in the scenarios (1 = Not at all frequently to 7 = Very frequently). The second item assessed the respondent’s view of acceptability of the communication (‘If I were [the nurse in this scenario], I would consider the perpetrator’s [perpetrator of the verbal abuse] verbal communication to be acceptable’) with a seven point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree). The next four items assessed the participant’s emotional response to the communication (e.g., ‘If I were [the nurse in this scenario], I would feel. 1 = Not at all depressed to 7 = Extremely depressed). In an effort to establish an overall measure of affective response to verbal abuse, a composite variable was computed based on the total score of the four affective responses (angry, depressed, upset, and bothered). Correlations among the four affective responses (across all scenarios) were all significant (p <.01) and ranged from 0.58 to 0.81, indicating moderate to high levels of common variance. A Cronbach’s alpha coefficient of 0.88 was calculated using the four items as part of a scale (again, across the scenarios) to provide some evidence that the composite measure was tapping a general level of emotional arousal.

The final four items of the ten item scale assessed the participant’s intended behavioural response to communication in terms of verbal abuse. The four behavioural intentions, identified within the study, could be positioned on a continuum from passive through to active. An entirely passive intended behavioural response to verbal abuse would be to ‘ignore the verbal abuser’ (Ignore). A slightly more active, but indirect option would be to ‘report the incident to your supervisor’. Rakos (1991) argued that assertive behaviour reflects the expression of one’s opinions and wishes directly, and for this reason Report was seen as an indirect response and next on the continuum from passive to active. A more active and direct responses to the verbal abuse would be to ‘tell the verbal abuser that the communication was not acceptable’ (Tell Them). This reflects the classic assertive response taught in social skills training. Finally, at the more active end of the continuum is an aggressive response ‘shout back at the verbal abuser’ (Shout Back). Although this response would be classified as aggressive rather than assertive, given the widespread culture of verbal abuse in the hospital setting, it was considered a likely, but active behavioural intention. Hence, it should be noted that behavioural intentions are operationalised as including both assertive and aggressive responses as the participant’s level of assertiveness may also predispose them to aggressive behaviour given the strong influence of hospital norms supporting such behaviour. Participants rated how likely they would be to use each of these responses in relation to the behaviour described in the scenario (1 = Not at all likely to 7 = Extremely likely). Participants were also given the opportunity to write their preferred response to the communication scenarios in case the seven point interval scales had not fully captured their response. These responses did not capture any different responses to verbal abuse. However, some of the comments supported the results, and these responses are referred to in the discussion section.

In an endeavour to circumvent priming effects, the Rathus Assertiveness Schedule (RAS) was placed at the end of the questionnaire, on a separate sheet of paper, so that the scenarios would be answered first. It was anticipated that this sequencing would also reduce the incidence of the participants guessing the hypotheses of the study and inadvertently affecting the results. The RAS has been shown to have moderate to high test-retest reliability (r = .78; p < .01) and split half reliability (r = .77; p < .01). Criterion validity in terms of behavioural responses in specific situations was satisfactory (r = .70; p < .01) (Rathus 1973). Although the RAS is more than 30 years old, recent psychometric tests have supported its continued use (Gustafson 1992). For each of the 30 items a rating scale of -3 to +3 is used. Seventeen items within the scale were reverse scored. Total scale scores range from -90 to + 90. A negative score reflected non assertiveness, whereas a positive score reflected assertiveness (Fischer & Corcoran 1994). Cronbach’s alpha for this scale in the present study was 0.87. These scores had to be converted to a dichotomous variable to allow for inclusion as a grouping variable in the ANOVA. Thus, participants with scores below zero were coded as non assertive and those with scores above zero were coded as assertive.

Analysis

Responses on each of the first three dependent variables (1. frequency of similar episodes, 2. acceptability of the episode, and 3. affective response to the episode) were analysed in separate 2 (Role: doctor, nurse) x 2 (Gender: male, female) x 2 (Situation: emergency, non emergency) x 2 (Assertiveness: assertive, non assertive) mixed within and between subjects ANOVAs. Role, Gender and Situation were within subjects factors, and Assertiveness was a between subjects factor. Dependent variables were the level of perceived acceptability of the verbal abuse, the composite affective reactions score, and each of the separate behavioural intentions.

Results

Data Screening

Data analysis was conducted using the SPSS 11 .0 for Windows software programme. Prior to analysis, all dependent measures were examined through the SPSS programme to assess accuracy of data entry, the fit between variable distributions, and the assumptions of analysis of variance (ANOVA). For this study, there were a total of 89 variables. Data screening revealed that the assumption of normality had been violated for a proportion of these variables. Indeed 23 (25.84%) of the variables were found to have a positive skewness statistic greater than ±1. As both positive and negative violations of normality were represented, data transformation was not considered a viable option. In view of this violation of the normality assumption of analysis of variance (ANOVA), the more stringent epsilon adjusted values were used to interpret within subjects effects. The Greenhouse- Geisser estimation was chosen as it produces a stronger adjustment, and thus, offers a more conservative result (Tabachnick & Fidell 2001). Significance levels were set at .05 for all statistical analysis. However, more conservative p values are reported.

Frequency of Verbal Abuse

Across all scenarios 61.43 per cent of the nurses indicated that verbal abuse was not frequent (scores 1.00-2.99), whereas 37.14 per cent reported moderately frequent verbal abuse (scores 3.00-5.00), and a further 1.43 per cent described the abuse as very frequent (scores 5.01-7.00). An analysis of the reported frequency of the participant’s experience of similar verbal abuse scenarios revealed a significant interaction between Role and Gender, F (1, 60) = 26.29, p < .001, η2 = .31. Furthermore, the participants reported that verbal abuse from male doctors was most frequent, with little difference in the frequency of abuse from female doctors and male and female nurses. Table 1 provides the means and standard deviations for each scenario.

Table 1
Means of Frequency of Abuse
Non Emergency Situation Emergency Situation
MD FD MN FM MD FD MN FM
3.512.872.422.552.992.482.282.34
(1.39)(1.33)(1.28)(1.31)(1.50)(1.23)(1.26)(1.28)

Notes:
a. MD = Male doctor, FD = Female doctor, MN = Male nurse, and FN = Female nurse.
b. Values in parentheses are the standard deviations of the means.

Additional analyses also revealed significant main effects for Role, F (1, 60) = 15.01, p < .001, η2 = 0.2; Gender F (1, 60) = 18.85, p < .001, η2 = 0.24; and Situation, F (1, 60) = 8.21, p < .01, η2 = 0.12. These main effects showed that verbal abuse was reported to be more frequently exhibited by males (M = 2.73, SE = .13) than females (M = 2.50, SE = .12), by doctors (M = 2.88, SE = .14) than nurses (M = 2.35, SE = .13), and in non emergency situations (M = 2.8, SE = .13) than in emergency situations (M = 2.44, SE = .14). A significant main effect of Assertiveness, F (1, 60) = 9.20, p < .01, η2 = .13 indicated that non assertive participants (M = 2.98, SE = .17) reported more frequent experience of similar verbal abuse scenarios than did assertive participants (M = 2.25, SE = .12).

Acceptability of Verbal Abuse

A significant interaction between Role and Situation, F (1, 60) = 5.51, p < .05, η2 = .08, and a main effect of Situation, F (1, 60) = 34.95, p < .001, η2 = .37, emerged in the analysis of the acceptability of the verbal abuse scenarios. Averaging over role, verbal abuse in non emergency situations (M = 1.50, SE = .09) was considered to be less acceptable than in emergency situations (M = 2.66, SE = .21). However, in an emergency situation participants found verbal abuse by a nurse (M = 2.50, SE = .21) slightly less acceptable than verbal abuse by a doctor (M = 2.82, SE = .22), whereas in a non emergency situation verbal abuse by a doctor or a nurse was considered similarly unacceptable (M = 1.46, SE = .11; and M = 1.56, SE = .12, respectively).

Affective Response to Verbal Abuse

Analysis of affective response to the verbal abuse scenarios demonstrated a significant three way interaction between Situation, Gender and Assertiveness, F (1, 60) = 12 .13, p >.001, η2 = .17, and two way interactions between Gender and Assertiveness, F (1, 60) = 3.87, p < .05, η2 = .06, and between Situation and Assertiveness, F (1, 60) = 6.47, p < .05, η2 = .10. Also subsumed under the interactions were significant main effects of Gender, F (1, 60) = 5.77, p < .05, η2 = .09; Situation, F (1, 60) = 23.67, p < .001, η2 = .28; and Assertiveness, F (1, 60) = 5.27, p < .05, η2 = .08.

Figure 1
Affective Response of Assertive and Non assertive Participants
Affective Response of Assertive and Non assertive Participants

Note: Higher numbers represent a more negative affect (angry, depressed, upset and bothered).
The vertical lines around the mean represent standard error of the mean.

The interaction between Situation, Gender, and Assertiveness is presented in Figure 1. Precisely, it is shown that overall, both assertive and non assertive respondents reported a stronger affective response to verbal abuse in a non emergency situation than in an emergency situation. Nevertheless, the effect was much larger for non assertive respondents. Furthermore, in a non emergency situation assertive respondents had a more negative response to verbal abuse from males than females, whereas the participants did not differ significantly in their responses across genders in emergency situations.

Behavioural Intentions

The data show a variety of reflections to verbal abuse. Overall, nurses reported that their most likely response to the verbal abuse was to Shout Back (M = 4.78, SE = .20), followed by Report (M = 4.16, SE = .17), then Ignore (M = 3.49, SE = .19). The least likely response was Tell Them (M = 2.12, SE = .14). This pattern of responses was reflected in a significant main effect for Response, F (1, 60) = 50.25, p < .001, η2 = .46.

Analysis of the behavioural intentions of participants in response to verbal abuse also revealed main effects for Assertiveness, F (1, 60) = 5.51, p < .05, η2 = .08, and for Situation, F (1, 60) = 45.86, p < .001, η2 = .43. The main effect for Situation was qualified by an interaction with Response, F (1, 60) = 3.76, p < .05, η2 = .06. A two way interaction between Assertiveness and Response F (1, 60) = 4.49, p < .01, η2 = .07 also emerged. Finally, analysis also demonstrated an interaction between Role and Response, F (1, 60) = 12 .69, p < .05, η2 = .18.

Figure 2
Assertive and Non assertive Participants’ Ratings to Verbal Abuse
Assertive and Non assertive Participants' Ratings to Verbal Abuse

The interaction between Assertiveness and Response is presented in Figure 2. It shows that the likelihood of each of the active behavioural responses (Report, Tell Them and Shout Back) was rated higher by assertive than non assertive participants, whereas the passive response of Ignore was rated as slightly more likely by non assertive than assertive participants. It is demonstrated in Figure 3 that across all response types, the likelihood of making a response (even to Ignore) was rated as higher in a non emergency situation than in an emergency situation. However, for the Tell Them response, there is little difference between the likelihood in non emergency and emergency situations. This effect may be driven by a floor effect in the likelihood of the Tell Them response, which was rated the least likely response.

Figure 3
Responses to Verbal Abuse in Emergency and Non emergency situations
Responses to Verbal Abuse in Emergency and Non emergency situations

Figure 4 illustrates the interaction between role and response. It is shown in Figure 4 that there is little difference in participants’ intended responses to verbal abuse from doctors and nurses in relation to intentions to tell the abuser that the communication was not acceptable or intending to shout back. However, participants were more likely to ignore the abuser when the abuse came from a doctor than a nurse, and more likely to report a nurse than a doctor.

Figure 4
Responses to Verbal Abuse From Doctors and Nurses
Responses to Verbal Abuse From Doctors and Nurses

Discussion

The present study investigated the relationship between role, gender, situational, and assertiveness variables and a range of reactions to scenarios describing verbal abuse in a hospital setting. The following discussion examines the findings in relation to each of the hypotheses that were made. The structure of this discussion, which is in two main sections, is based around the literature that underpinned the predictions. Firstly, although not a specific hypothesis of the current study, the prevalence of verbal abuse experienced by nurses in this sample is examined. Next, each of the main variables of interest, derived from the major theoretical perspectives, and which were elucidated in the introduction in the categories of gender, role, situation, and assertiveness, are discussed. This discussion is in relation to the nurses’ affective, cognitive and behavioural responses to verbal abuse as were reported in the present study.

Prevalence of Verbal Abuse

Across all scenarios, 38.6 per cent of practising nurses in this sample described their experience of verbal abuse similar to that described in the scenarios as moderately frequent or very frequent. It should be noted that the question specifically asked about experience as involving being a target or witness of such behaviour in order to gauge a general level of the occurrence of the behaviour. The level of reported frequency in this study is similar to the self reports of workplace violence that were identified by the Queensland Nurse’s Union (Hegney, et al. 2003). More specifically, the most frequent verbal abuse was reportedly received from male doctors, with lower reported frequency of abuse from female doctors as well as male and female nurses. This finding supports previous literature, which indicated that doctors are a more frequent source of verbal abuse than nurses (Begany 1995, Bruder 2001, Cook, et al. 2001, Buback 2004). In addition, a feminist analysis of verbal abuse implicated power inequality between genders as the source of the behaviour (Bruder 2001). However, as the question asked participants to reflect on actual rather than relative frequencies in their work experiences, lower frequencies for female doctors and male nurses may reflect more limited experiences of interacting with members of these groups (and, therefore, fewer opportunities for abusive behaviour). Assuming that exposure with male and female doctors, and nurses is common, the results indicate that, in general, verbal abuse is more frequently received from doctors than nurses. It is also important to note that the study did not involve male nurses as participants (as receivers of verbal abuse behaviour). Therefore, it is unknown whether male nurses also would have reported similar levels of abuse from male doctors. An important direction for future research is to examine whether male nurses report as much verbal abuse from male doctors as female nurses. If so, the role based (status) interaction may be the stronger determinant of the behaviour than the gender based interaction.

The study results also indicated that verbal abuse enacted by doctors or nurses of either gender was more frequent during non emergency situations than emergency situations. It could be argued that this result is contrary to what might be expected, in that stress is cited as a common reason for the occurrence of verbal abuse. Alternatively, the remembered incidence of verbal abuse may have been affected by the perceived acceptability of abuse in particular situations. Participants considered verbal abuse much less acceptable in non emergency than emergency situations. Therefore, participants may fail to recall incidents of verbally abusive behaviour similar to that described in the scenarios that occurred in emergency situations, because the behaviour stands out less in emergency situations, therefore, receiving less attention and being less likely to be encoded into memory.

Assertive respondents also reported lower frequency of similar scenarios, suggesting that assertiveness training may be effective in reducing the actual incidence of verbal abuse. Alternatively, for the relationship between situation and frequency, it may be that instances of abusive communication may have received less processing by assertive participants, who on average, were also less affected emotionally by verbal abuse than non assertive participants.

Gender

It was hypothesised in H1 that verbal abuse would be considered less acceptable and more upsetting when it was enacted by a member of the same gender (female) than the opposite gender (male). This was expected as it was theorised that aggressive communication is less aligned with a female social role, and according to Social Identity theory, females will be affiliated with one another as members of an ingroup, and are, therefore, expected to possess more positive qualities than outgroup members (males) (Worchel, et al. 2000). Based on social role theory and gender power inequality, it was also predicted that the participant’s behavioural intentions towards the scenario would be less assertive or aggressive towards a male than a female. It was found that the gender of the abuser interacted with the situation and the nurse’s assertiveness in terms of the affective response to abuse, such that there was a tendency in a non emergency situation for assertive participants to react more negatively to verbal abuse from a male than from a female. Perhaps the perception that men misuse their higher status and power related to gender may lead to higher affective arousal in assertive participants. However, apart from this finding, the hypothesis outlined in H1 garnered limited support in that gender seems to have little effect on the acceptability of verbal abuse, or on the affective or behavioural responses to abuse.

Role

It was predicted in hypothesis H2 that it would be less acceptable and more upsetting to be verbally abused by another nurse than by a doctor. There was a tendency for verbal abuse to be considered less acceptable from a nurse than a doctor, but only in an emergency situation. This finding may lend some support to Social Identity theory, as it is argued that group identification is heightened under threat conditions (Tajfel & Turner 2001). Alternatively, nurses may be responding to social rules governing appropriate social role behaviours. In short it may be acceptable for a doctor to become verbally abusive under the stress of an emergency situation. However, there was no indication that participants found the verbal abuse from a nurse more upsetting than from a doctor in any situation. Therefore, the component of H2 that related to cognitive and affective responses received only limited support.

In terms of the participants’ behavioural responses to verbal abuse contained in H2, there was some support for the hypothesis that participants would use a more assertive response with a nurse than a doctor. Participants were more likely to use the most passive response of ignoring the abuser when they were a doctor rather than a nurse, and more likely to respond by reporting to their supervisor abuse from a nurse than abuse from a doctor. As the intended behaviour ‘ignore the verbal abuser’ was considered to be a passive behaviour, and the behaviour ‘report to your supervisor’ more assertive (though still relatively passive), these results could be seen to support the expectation that assertive behaviour is less likely to be used when a status difference (based on role) exists. Thus, behavioural responses to abuse are influenced by the social rules that shape what people consider appropriate behaviour when responding to someone of higher status, such as a nurse responding to a doctor (Wilson, et al. 2001). However, there was little indication that the more assertive response of telling the verbal abuser that the communication was not appropriate, or the aggressive response of shouting back, were affected by the status of the person delivering the abuse.

Situation

Predictions contained in the hypothesis H3 related to the influence of the situation on reactions to verbal abuse. These contentions were based on Attribution theory in speculating that emergency situations would provide an ‘external reason’ for verbal abuse to be more acceptable, less upsetting, and less amenable to assertive intervention. The conditions included in H3 were generally supported by the results of this study. Nurses rated verbal abuse as more acceptable and reported lower levels of emotional arousal when it occurred during an emergency, rather than a non emergency situation. Thus, as described by Heider’s (1958) Attribution theory, it could be argued that the participants made external attributions about the actor of verbal abuse, acting to provide an ‘excuse’ for the behaviour, and thus, take the abuse ‘less personally’. Also, for each of the intentioned behavioural responses to verbal abuse there was a tendency for participants to endorse a higher likelihood of responding in a non emergency situation, than in an emergency situation. Interestingly, this was even the case for the intention to ignore the abuser. This observation may suggest that participants interpreted ignoring the abuser as an active act, and perhaps found it difficult to picture themselves responding in any way to verbal abuse in an emergency situation, focusing instead on the job at hand. Two comments from the participants further illustrated these findings. First ‘Get on with the job — emergency situations do not lend themselves to discussion on etiquette’ is a salient reaction to the abuse and situation context. And the second comment ‘doctors often cope with stress via emotional outbursts, even though to ‘new’ nurses this is unacceptable, the more you are in the profession of nursing the more apt you become to accept ‘outbursts’ and move on’ is a further illustration. These comments also suggest that one of the aims of assertiveness training, the ability to not take verbal abuse to heart, might indeed, assist nurses to cope with a climate of verbal abuse.

Assertiveness

Significant support was not found for hypothesis H4. This speculative arrangement contended that assertive participants would consider verbal abuse as less acceptable than non assertive respondents. The hypothesis H4 also predicted that assertive respondents would report less affective arousal to verbal abuse than non assertive respondents, due to common coping strategies associated with assertiveness (e.g., see the behaviour as a problem with the actor of the abuse and not let it affect their personal self esteem). This component of the H4 was supported in that non assertive nurses appeared to experience higher levels of negative emotional arousal than their assertive counterparts, particularly in non emergency situations. It could be argued that non assertive participants may be more likely than assertive participants to feel that they are the cause of the actor’s verbal abuse in a non emergency situation, which may add to their affective arousal.

Behavioural Intentions

The predictions made in the hypotheses H5 and H6 regarding the likely intended behavioural responses of nurses in different scenarios were predicated on the assumption that the behavioural responses were on a continuum with extremes of passive to active reactions. Specifically, to ignore the abuser was considered to be the most passive response, followed by reporting the incident to a supervisor. Telling the abuser that the communication was unacceptable was considered an active, assertive response. Alternatively, shouting back at the abuse was categorised as an active, aggressive response. The overall pattern of results for the four behavioural responses suggests that participants may have viewed the assertiveness, or appropriateness, of the responses differently to the theoretical conception. Averaging across all scenarios, the behaviour with the most highly endorsed likelihood was shouting back at the abuser, whereas telling the abuser that the communication was unacceptable was the least likely response. It is probable that the nurses’ behavioural intentions were shaped by the strong organisational norms supporting aggressive behaviour or retaliatory verbal abuse. Smith, et al. (1996) reported that nurses described their working environment as hostile, where they experienced disrespectful treatment, and that a lack of support was commonplace. In such a context, aggressive responses to workplace violence may be considered appropriate. Assertive responses, which emphasise the needs of the individual, such as directly telling another that their communication is not acceptable, may be considered inappropriate within a hospital setting, where the needs of the patients should take priority over individual feelings. In contrast, a system of reporting behaviour to a supervisor fits in well with the hierarchical organisational structure and public service culture of hospitals.

There was some support for H5 in that assertive participants would be more able to employ active behaviours, such as expressing their rights, which are thought to aid them in confrontational situations (Alberti & Emmons 1982). For the more active behavioural responses (i.e., Report, Tell Them & Shout Back) the assertive participants endorsed higher likelihood than did non assertive participants. However, the largest difference between assertive and non assertive participants’ likelihood ratings was for the Report behaviour, with an assertive participant more likely to respond to abuse by reporting it to a supervisor than a non assertive participant. There was no significant evidence that assertive participants were selectively choosing the more assertive or aggressive behavioural responses and that non assertive participants were choosing the passive responses as had been predicted in H6. Assertive participants endorsed higher likelihoods overall, and assertive and non assertive participants seemed similarly reluctant to tell the person their communication was inappropriate, and similarly ready to shout back at verbal abuse. Whilst assertiveness training might encourage a direct personal response (i.e., Tell Them) to a verbal abuser, it appears that even assertive nurses shaped the style of their behavioural intention to fit the organisational culture.

In conjunction with the frequency data, these results suggest that there may be some benefit to individual assertiveness training for nurses. The results lend further justification to previous researchers’ arguments that assertiveness training can assist nurses to cope with the stress of verbal abuse attacks (McIntyre, et al. 1984, Cook, et al. 2001, Mimura & Griffiths 2003). Assertive nurses appeared to experience fewer incidences of similar verbal abuse, were less affected emotionally by such abuse, and more likely to respond to abuse in an assertive manner. However, because the most likely response for all participants could be constituted as aggressive, the results also suggest that change is necessary at an organisational normative level as a complement individual assertiveness training. This would enable assertive behaviours to be contextually appropriate, addressing one of the major criticisms of assertiveness training (Wilson, et al. 2001).

Boundary Conditions and Directions for Future Research

The study findings are bounded by the employed methodology. Measurement of all study variables in a single questionnaire is problematic as the single source nature of the data raises the possibility of common method variance. In addition, placement of the assertiveness schedule (RAS) at the end of the questionnaire could have resulted in priming in the opposite direction anticipated, which may have confounded the results. Further, scenario studies measure reported intentions or the participant’s estimate of their likely responses in such scenarios, not their actual responses to experienced verbal abuse. And as it is difficult to test a design as complex as that employed in this study in a field study, an experimental study involving simulated behaviour in a laboratory setting may prove useful as a means of replicating the reported findings. Such a study could consider using observer ratings of behaviours exhibited and other sources of data on a participant’s assertiveness. An additional feature for consideration relates to the use of a fairly crude measure of affective response to verbal abuse developed for use in the present research (a composite of the four affective responses — angry, depressed, upset, and bothered). Future research could improve the measure of affective response and include predictions about the type and strength of various types of emotions in relation to the situational variables studied.

Conclusion

Both the occurrence of, and response to, verbal abuse appeared to be under strong situational control. Verbal abuse was more frequent, considered less acceptable and elicited more affective arousal in non emergency than in emergency situations, and the situation influenced the nurses’ willingness to engage in a behavioural response to the verbal abuse. The results suggest that situation is a stronger determinant of the occurrence of, and response to, verbal abuse than gender or role of the perpetrator of verbal abuse. Assertive individual’s experiences of verbal abuse also seemed to be different to those of non assertive individuals. More assertive nurses recalled experiencing fewer instances of verbal abuse, and responded differently to abuse than did less assertive nurses. Assertive nurses reported responding to verbal abuse with lower levels of negative emotional arousal and with more assertive behavioural intentions than non assertive nurses. Assertive nurses’ affective responses seemed to be less affected by situation than those of non assertive nurses, though there was some evidence for an influence of gender on assertive nurses’ emotional reactions to abuse. Overall, role and gender had more limited impact on experiences of verbal abuse. Male doctors were the most frequently cited perpetrators of verbal abuse towards nurses, but although role influenced the acceptability of verbal abuse and gender the affective response to abuse, these effects were limited to specific situations. These effects suggest that nurses’ attitudes and response are to some degree influenced by their social identity, but perhaps under greater situational control. The finding that nurses were more likely to use a passive response with a doctor than a nurse, and to use a more assertive response with a fellow nurse, also suggests a strong influence of social rules governing appropriate and acceptable forms of behaviour by doctors and nurses in different situations.

Within this study it would appear that the occurrence of vertical violence, in the form of verbal abuse, was more common and had the potential for greater impact than horizontal violence. However, horizontal violence behaviours tend to be more covert than verbal abuse. For example, faultfinding, backstabbing, and subtle sabotage are subtle responses to perversion (Smith, et al. 1996). Therefore, it would be interesting to ascertain the prevalence, tactics and psychological impact related to these types of workplace violence. Another area of interest could include the examination of respondent self esteem and other individual difference variables besides assertiveness. Finally, the study focused specifically on verbal abuse of nurses in a small group of hospital employees. For future studies it could be beneficial to use larger samples and organisations other than hospitals to test whether similar results occur in other work contexts, in particular, whether situation has as strong an impact in less structured workplaces.

The present research found evidence that cognitive, affective and behavioural attitudes towards verbal abuse scenarios varied as a function of the situation in which it occurred, whilst also being affected by the gender and role of the actor. As verbal abuse contributes to dissatisfaction and turnover, managers across a wide range of settings in addition to healthcare environments will be challenged to recognise the emotional impact of such behaviours and how to implement effective strategies of prevention and management. Intervention strategies within hospital settings need to address the relatively high incidence of verbal abuse directed at nurses by male doctors.

Indeed, in all workplaces managers should be aware that verbal abuse may be systematically influenced by social role, which has serious implications for organisations where status differences are inextricably linked with gender and role. Moreover, managers should be particularly cognisant of the strong influences of situation on employee responses to verbal abuse in a wide range of organisational settings. Finally, assertiveness training may help individuals to cope with verbal abuse, but it is important to recognise that organisational level interventions such as codes of conduct, grievance processes, and education strategies that specifically address workplace violence will need to be adopted to deal with the counter productive behaviour of the perpetrator’s verbal abuse.

Author

Dr Angela Martin is a lecturer in management at the University of Tasmania. Her PhD was in the area of employee adjustment during organisational change and her research interests include employee mental health and wellbeing and organisational communication.

Email: Angela.Martin@utas.edu.au

Ms Cynthia Gray is a registered nurse and a Master of Psychology student at the University of Southern Queensland.

Email: Andrea.Adam@utas.edu.au

Dr Andrea Adam is a lecturer and researcher in psychology at the University of Tasmania.

Email: Andrea.Adam@utas.edu.au

Acknowledgement

History of the manuscript. This article is based on research conducted for an honours dissertation by the second author, but has been prepared for publication by the first author, who was the supervisor of the work, with assistance from the third author. An abridged version of this paper (10 pages), focussing only on main the effects was presented at the 2005 Australia New Zealand Academy of Management conference.

References

Alberti, R. E., & Emmons, M. L. (1982). A guide to assertive living, your perfect right (4th ed.). Los Angeles: Impact Publishers.

Antari-Otong, D. (2001). Creative stress-management techniques for self-renewal. Dermatology Nursing, 13(1), 31-43.

Arway, A. G. (2002). Causal factors of violence in the workplace: A human resource professional’s perspective. In M. Gill, B. Fisher & V. Bowie (Eds.), Violence at work, causes, patterns and prevention (41-58). Oregon, United States of America: Willan Publishing.

Ashkanasy, N. M., & Gallois, C. (1994). Leader attributions and evaluations: Effects of locus of control, supervisory control, and task control. Organisational Behaviour and Human Decision Processes, 59(1), 27-50.

Baron, R. A. (1985). Reducing organizational conflict: The role of attributions. Journal of Applied Psychology, 70(3), 434-441.

Barron, O. (2002). Why workplace bullying and violence are different: Protecting employees from both. In M. Gill, B. Fisher & V. Bowie (Eds.), Violence at work, causes, patterns and prevention (151-164). Oregon, United States of America: Willan Publishing.

Begany, T. (1995). Do you get the respect you deserve? RN, 58(5), 32-34.

Bowie, V. (2002). Defining violence at work: A new typology. In M. Gill, B. Fisher & V. Bowie (Eds.), Violence at work, causes, patterns and prevention (1-20). Oregon, United States of America: Willan Publishing.

Bruder, P. (2001). Verbal abuse of female nurses: An American medical form of gender apartheid? Hospital Topics, 79(4), 30-35.

Buback, D. (2004). Assertiveness training to prevent verbal abuse in the OR. AORN Journal, 79(6), 148-164.

Burke, R. J. (1993). Organisational level interventions to reduce occupations stressors. Work and Stress, 7(11), 77-87.

Bush, H. A., & Gilliland, M. S. (1995). Caring for the nurse self: Verbal abuse as a case in point. Journal of Nursing Care Quality, 9(4), 55-62.

Cook, J. K., Green, M., & Topp, R.V. (2001). Exploring the impact of physician verbal abuse on perioperative nurses. AORN Journal, 74(3), 317-29.

Cox, H.C. (1991). Verbal abuse nationwide, part 1: Oppressed group behaviour. Nursing Management, 22(2), 32-36.

Cox, H. C. (1994). Excising verbal abuse. Today’s OR Nurse, 16(4), 38-40.

Cusack, S. (2000). Workplace bullying: Icebergs in sight, soundings needed. The Lancet, 356(9248), 2118.

De Frank, R, S., & Cooper, C. L. (1987). Worksite stress management interventions: Their effectiveness and conceptualisations. Journal of Managerial Psychology, 2(1), 4-9.

Diekman, A. B., & Eagly, A. H. (2000). Stereotypes of dynamic constructs: Women and men of the past, present, and future. Personality and Social Psychology Bulletin, 26(10), 1171-1188.

Douglas, S. C., & Martinko, M. J. (2001). Exploring the role of individual differences in the prediction of workplace aggression. Journal of Applied Psychology, 86(4), 547-559.

Duffy, E. (1995). Horizontal violence: A conundrum for nursing. Royal College of Nursing, Australia, 2(2), 5-17.

Dunn, H. (2003). Horizontal violence among nurses in the operating room. AORN Journal, 78(6), 977-987.

Durkin, N., & Wilson, C. (1999). Simple steps to keep yourself safe. Home Healthcare Nurse, 17(7), 430-435.

Eagly, A. H. (1987). Sex differences in social behaviour: A social-role interpretation. Hillsdale, New Jersey: Erlbaum.

Elgin, S. H. (1980). The gentle art of self-defence. New Jersey: Prentice Hall.

Farrell, G. A. (1999). Aggression in clinical settings: Nurses’ views - a follow up study. Journal of Advanced Nursing, 29(3), 532-541.

Fischer, J., & Corcoran, K. (1994). Measures for clinical practice: A source book, Volume 2, Adults (2nd ed.). New York: The Free Press.

Fiske, S. T., & Taylor, S. E. (1984). Social cognition. New York: McGraw Hill.

Fox, S., & Spector, P. (2005). Counterproductive work behaviour. Washington, DC: American Psychological Association.

Freire, P. (1970). Pedagogy of the oppressed. New York: The Seabury Press.

Gill, M., Fisher, B., & Bowie, V. (2002). Violence at work, causes, patterns and prevention. Oregon, United States of America: Willan Publishing.

Gustafson, R. (1992). A Swedish psychometric test of the Rathus assertiveness schedule. Psychological Reports, 71(2), 479-482.

Hegney, D., Plank, A., & Parker, V. (2003). Workplace violence in nursing in Queensland, Australia: A self-reported study. International Journal of Nursing Practice, 9 (12), 261-268.

Heider, F. (1958). The psychology of interpersonal relations. New Jersey: Lawrence Erlbaum.

Inglehart, R., & Norris, P. (2003). Rising tide, gender equality and cultural change around the world. Cambridge: Cambridge University Press.

Keashly, L. (2001). Interpersonal and systematic aspects of emotional abuse at work: The target’s perspective. Violence and Victims, 16(3), 233-268.

Lopez, D. (1993). Surviving verbal abuse: A win-win game plan. Nursing Management, 24(5), 80-82.

McCall, E. (1996). Horizontal violence in nursing: The continuing silence. The Lamp, 53 (3), 28-31.

McIntyre, T. J., Jeffrey, D. B., & McIntyre, S. L. (1984). Assertion training: The effectiveness of a comprehensive cognitive – behavioural treatment package with professional nurses. Behaviour Research & Therapy, 22(3), 311-318.

Milstead, J. A. (1996). Conflict management and negotiation. Orthopaedic Nursing, 15 (2), 39-47.

Mimura, C., & Griffiths, P. (2003). The effectiveness of current approaches to workplace stress management in the nursing profession: An evidence based literature review. Occupational and Environmental Medicine, 60(1), 10-16.

Muff, J. (1982). Socialization, sexism and stereotyping, women’s issues in nursing. Prospect Heights, Illinois: Waveland Press.

Numerof, R. E. (1978). Assertiveness training for nurses in a general hospital. Health and Social Work, 3(1), 80-101.

Paterson, B., McCornish, A. G., & Bradley, P. (1999). Violence at work. Nursing Standard, 13(2), 43-46.

Perrone, S. (1999). Violence in the workplace: Australian Institute of Criminology research and public policy series No. 22. Canberra, Australia: Australian Institute of Criminology.

Pokalo, C. L. (1991). Verbal abuse and sexual harassment in the OR. Today’s OR Nurse, 13(9), 4-7.

Queensland Government. (2001). Workplace Bullying issues paper. Queensland Workplace Bullying Taskforce. Division of Workplace Health and Safety, 1-63.

Rakos, R. F. (1991). Assertive behaviour: Theory, research and training. London: Routledge.

Rakos, R. F. (1997). Asserting and confronting. In O. D. W. Hargie (Ed.), The handbook of communication skills (2nd ed., 289-315). London: Routledge.

Rathus, S. A. (1973). A 30-item schedule for assessing assertive behaviour. Behaviour Therapy, 4(1), 398-400.

Roberts, S. J. (1983). Oppressed group behaviour: Implications for nursing. Advances in Nursing Science, 5(3), 21-30.

Roberts, S. J. (1997). Nurse executives in the 1990s: Empowered or oppressed? Nursing Administration Quarterly, 22(1), 64-72.

Roberts, S. J. (2000). Development of a positive professional identity: Liberating oneself from the oppressor within. Advances in Nursing Science, 22(4), 71-84.

Simms, C. (2000). Stopping the word war. Nursing Management, 31(9), 65-72.

Smith, L. S. (1997). Collaboration: The key to assertive communications. MedSurg Nursing, 6(6), 373-375.

Smith, M., Droppleman, P., & Thomas, S. P. (1996). Under assault: The experience of work-related anger in female registered nurses. Nursing Forum, 31(1), 22-34.

Sofield, L., & Salmond, S.W. (2003). Workplace violence: A focus on verbal abuse and intent to leave the organization. Orthopaedic Nursing, 22(4), 274-284.

Spreitzer, G. M. (1997). Toward a common ground in defining empowerment. In W. A. Pasmore, & R. W. Woodman (Eds.), Research in organizational change and development (31-62). Greenwick, CT: JAI Press.

Tabachnick, B. G., & Fidell, L. S. (2001). Computer-assisted research design and analysis. Boston: Allyn and Bacon.

Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (33-47). Monterey, CA: Brooks/Cole.

Tajfel, H., & Turner, J. (2001). An integrative theory of intergroup conflict. In M. A. Hogg & D. Abrams (Eds.), Intergroup relations: Essential readings (94-109). Philadelphia: Psychology Press.

Taylor, B. (2001). Identifying and transforming dysfunctional nurse-nurse relationships through reflective practice and action research. International Journal of Nursing Practice, 7 (6), 406-413.

Vogt, J., Cox, J., Velthouse, B., & Thames, B. (1983). Retaining professional nurses: A planned process. St Louis: C.V. Mosby.

Wilson, K. L., Lizzio, A. J., & Zauner, S. (2001). Social rules for managing attempted interpersonal domination in the workplace: Influence of status and gender. Sex Roles, 44 (3-4), 129-154.

Woodward, E. K., & House, B. M. (1997). Nurse-physician communication: Women and men at work. Orthopaedic Nursing, 16(1), 39-46.

Worchel, S., Cooper, J., Goethals, G. R., & Olson, J. M. (2000). Social psychology. Belmont, California: Wadsworth.