Exploring Workplace Bullying in an Emergency Service Organisation in the UK

The data reported in this paper is drawn from a study of workplace bullying in an Emergency Service Organisation (ESO) in the United Kingdom. This ESO is dynamic and well ordered and a key characteristic of this organisation is that it is service driven. The most important role that many ESO members play is to save life and to ensure that people live in a safe environment. The ESO is also highly structured, hierarchical and power based, with a very strong discipline code. ESO staff are predominantly white male, with a high expectation that they work as a group. The management is highly authoritarian and operates in a command and control relationship between those that occupy a higher rank and those of lower rank. This is an exploratory research, and the data reported in this paper was drawn from a total of 452 people who responded to a questionnaire study, thereby achieving 25% response rate. The primary aim of the research reported in this paper is to test for significant differences in the kinds of bullying behaviours employees are exposed to in ESO. The study explored the different types of workplace bullying experienced. The methodology of the study incorporated online questionnaires and a postal survey using a single instrument, the Negative Acts Questionnaire, Revised (NAQ-R). A factor analysis and Wilcoxon Mann-Whitney Test was carried out on these three types of bullying and some of the demographic factors such as gender, sexual orientation, age, rank, length of service in the organisation, ethnicity and disability. The results were designated personal bullying, administrative bullying and social exclusion. It was discovered that employees experienced different types of bullying and that bullying was part of the culture of the organisation. The results from this study indicate that gender, sexual orientation, age, occupational group, length of service in the organisation, ethnicity and disability all play significant roles in the kinds of bullying behaviours to which employees are exposed.


Introduction
There has been an unprecedented growth in the academic literature on workplace bullying in the last fifteen years in which many of the arguments have highlighted the increasing need for organisations to address the problems caused by bullying in the workplace.Existing literature has shown that there are various approaches to studying bullying in the workplace, including a debate on how to define workplace bullying (Lewis & Sheehan, 2003;Sheehan, 2006).To date, there has been no general agreement on the definition of workplace bullying.However, even with several conceptual and methodological differences across various studies, there has been a growing convergence of definitions of workplace bullying in recent years by researchers such as Einarsen, Hoel, Zapf and Cooper (2003).For instance, in Europe, a group of researchers defined bullying at work as Harassing, offending, socially excluding someone or negatively affecting someone's work tasks.In order for the label bullying to be applied to a particular activity, interaction or process, it has to occur repeatedly and regularly for instance (weekly) and over a period of at least 6 months (Einarsen et al., 2003, p.15).
Workplace bullying may be seen, therefore, as an escalating process in which the person confronted becomes the target of systematic negative social acts (Notelaers et al., 2006).That is, bullying involves negative or hostile behaviours which occur regularly and repeatedly and over time, rather than being an isolated or single incident.Thus, this definition of workplace bullying tends to emphasise persistency and duration of exposure (Einarsen et al., 2006), which according to Hoel and Beale (2006) provides a distinction between workplace bullying and conflict.Leymann (1996) stressed that in some cases, incidences of bullying may arise as a result of unresolved conflicts.Einarsen et al. (2003) emphasised that the roles and the dynamics of the conflict escalation process were central to an understanding of the bullying process.There may be a severe incident in which the behaviour induces a negative effect such that it needs not to be repeated for the effect to remain (Zapf & Einarsen, 2003).In particular, some treatments, such as having work tasks taken away, may only happen once, but the consequences of the acts are experienced daily.Thus bullying consists of psychologically aggressive and hostile acts that are perceived by the subjects in a negative way (Salin, 2003).Bullying ought to be seen as a social process in which the impact on the person experiencing workplace bullying is of primary importance.
Whilst studies on workplace bullying provide an insight to the types, problems and the causes of workplace bullying, some of the confusion about defining workplace bullying according to Sheehan and Barker (1998) often lies with methodological choice.Two distinct research methodologies tend to have been used to investigate the phenomena of workplace bullying.The first is the population survey of working employees (includes both bullied and non-bullied).The second comprises studies of victims of bullying, often within a clinical setting (Leymann, 1996;Vartia, 2001).One of the tools often used in questionnaire studies is the Negative Acts Questionnaire revised edition (NAQ-R) developed by Einarsen and Hoel (2001).An adapted (NAQ-R) was used in the study from which this paper was drawn.
An argument relating to power imbalance has also been emphasised.It has often been argued that the targets or victims of bullying cannot defend themselves on an equal basis (Vartia, 2003;Salin, 2003).Within the parameters of this argument, workplace bullying has been seen as involving a 'victim-perpetrator' dimension (Salin, 2003), especially when the victim or target has been subjected to negative behaviours on a scale whereby he or she feels inferior in defending him or herself in the actual situation (Salin, 2003;Hoel, 2006;Sheehan, 2006;Vartia, 2003).Power imbalance can be in different forms, such as the formal power differences and abusive supervision sometimes found in a highly structured organisation with ranks and grades (Archer, 1999); or in informal social groups (Salin, 2003).The imbalance of power according to Fox and Spector (2005) often mirrors the formal power structures of the organisation in which someone has been on the receiving end of negative acts from a person in a superior organisational position.
The types of bullying behaviours are often defined as a combination of various factors, especially those based on the experiences and perceptions of those who have been exposed to workplace bullying.Several researchers have used the victims' accounts as a basis to indicate and identify the types of bullying behaviours reported (Ayoko, Callan & Hartel, 2003).Several types of workplace bullying behaviours have been identified, such as the withholding of information (Hoel & Cooper, 2000;Salin, 2004), verbal abuse (Rayner et al., 2003), insults and excessive teasing (Hoel et al., 2001), sexual harassment (Hoel & Cooper, 2003;Mikkelsen & Einarsen, 2001), and aggression (Bjorkqvist, Osterman & Hjelt-Back, 1994;Cowie, Naylor, Rivers, Smith & Pereira, 2002;Lee and Brotheridge, 2006).Determining the differences among the types of workplace bullying experienced revolves round various issues.To some, the type of workplace bullying experienced could depend on the tasks performed in the organisation or the different levels and positions occupied by the employee(s) within the organisation (Hoel & Cooper, 2000).For example, the roles performed by supervisors or managers and subordinates can be a factor that impacts upon the type of bullying they experience.In a study carried out by Paananen and Vartia (1991) cited in (Vartia 2003, pp.23), the over-assigning and the oversimplification of tasks were reported to be the type of bullying acts mostly used by supervisors, while threats or acts of physical violence were most often used by subordinates.In a more recent study carried out by Hoel and Cooper (2000), some managers reported that they had been exposed to 'unmanageable workloads' and 'unreasonable deadlines', whereas other workers reported that they had been exposed to 'insults or offensive remarks' and 'excessive teasing'.These empirical findings point to a relationship between position, tasks and duties performed in the organisation and the types of bullying acts experienced by the employees.
That is, in the early phases of workplace bullying, victims are typically subjected to aggressive behaviours that are difficult to pinpoint because they are often indirect and discreet (Vartia, 2003).However, if care is not taken, indirect aggressive behaviours can escalate to more direct, aggressive acts such as verbal or physical abuse (Einarsen, 2006;Leymann, 1996).Furthermore, workplace bullying has been reported as a special type of aggressive behaviour (Bjorkqvist et al., 1994).According to Neuman (2004), all acts of bullying begin with a single act of aggression, so anything that increases the likelihood of aggression may serve to increase the likelihood of bullying.Once a cycle of aggression between two parties begins, the process will continue, escalate and result in negative responses towards others (Andersson & Pearson, 1999;Cowie et al., 2002).Workplace aggression has been classified into direct and indirect aggressive behaviours (Bjorkqvist et al., 1994;Neuman, 2000).As suggested by Einarsen et al. (1998), in the early stages of bullying, perpetrators are most likely to engage in behaviours that are difficult to pinpoint because they are very indirect and discreet; later on they move to more direct aggressive acts.Direct aggression is the most common form of aggression used by students in the school environment (Olweus, 2004), while in the work environment, indirect aggression is most common (Vartia, 2003).Indirect aggression comes in different forms such as exposure to false accusations, rumours, gossip, malicious stories, belittlement of opinions and limitation of workers' opportunities to express their opinions (Bjorkqvist et al., 1994;Vartia, 2003).According to Lee and Brotheridge (2006), the most common form of aggression is verbal-passive indirect aggression, such as failing to transmit information needed by the target.Some of the aforementioned researchers named above suggest that bullying tactics include direct and indirect, and active and passive aggression actions.A less reported type of workplace bullying is sexual harassment.There is still ongoing debate concerning whether or not sexual harassment should be seen as another type of workplace bullying (Sheehan, 2006).Sexual harassment, according to McMahon (2000), is any conduct related to sex and sexual orientation which is unwanted by the recipient.Most of the research on workplace bullying has reported fewer accounts of sexual harassment as a type of workplace bullying.For example, in a study carried out by Piirainen et al. (2000) in Finland, 1.5% of the employees have been exposed to at least occasional sexual harassment; while Mikkelsen and Einarsen (2001) advised that 0.5-3.9% of the Danish respondents have been exposed to occasional sexual harassment.
Based on these typologies, it can be concluded that workplace bullying involves deliberate, hurtful and repeated mistreatment of employees in a work environment (Neuman, 2004).Although workplace bullying can occur among co-workers or be directed by subordinates against superiors, or superiors against subordinates (Neuman, 2004), one of the most reported reasons behind it, according to Lee (2002), is the abuse of power by superiors against subordinates.It is important to note that people hold legitimate and formal power, granted by the organisation, and informal power to establish superior-subordinate relationships (Palmer & Hardy, 2002;Vartia, 2003).For instance, The Chartered Institute of Personnel and Development (CIPD) (2004) in their UK study reported that in the manufacturing and production industry, 38% of the respondents reported having been bullied by line managers, and 39% by peer colleagues.This is similar to those of Einarsen and Raknes (1997) on male industrial workers in Bergen. McIvor (2007) also reported that workplace bullying is common in workplaces dominated by tough males and in institutions such as the police.Heol and Cooper (2000) study in the UK also reported 16.2% prevalence rate in the prison service.While Vatia (2003) research on work environment, well being and health using 949 municipal employees, 896 prison officers and 5,432 hospital employees in Finland reported that bullying is most common in prisons with prevalence rate of 20.1%, followed by municipal institutions 10.1% and hospital 5%.Similarly, Eriksen and Einarsen (2004) reported workplace bullying to be going in health service organisation in their Norwegian Union of Health and Social Workers study.They reported a prevalence rate of 4.3% for female respondents and 10.3% for male respondents.McCarthy et al. (2003) study on 100 staff members of a tertiary institution in Australia reported that 80% have experienced workplace bullying in the last 1 month.This indicates that bullying has a significance influence in the working life of the Australian employees.Furthermore, workplace bullying has been reported to be going on in the voluntary sectors.For instance, CIPD (2004) in their study reported that 33% were bullied by their line managers and 28% by peer colleagues in the voluntary sector.While in the non-profit sector, 10% were bullied by their line managers and 7% were bullied by their subordinates respectively.All these aforementioned researchers reported a case of a tough environment where harsh humour appeared to be part of everyday life and the accepted culture as factors that can enhance the emergence of workplace bullying.However, making comparisons between all these studies should be treated with some caution because workplace bullying is measured by either subjective or objective methods.The subjectivity and objectivity approaches play an important part in the development of research on workplace bullying.The differences in definitions, sample size, context, strategies and methods used may have affected the prevalence rates as reported by different researchers.It is therefore difficult to make international comparison.The use of a standard measure has suggested by Einarsen and Hoel (2001) would enable international convergence towards research on workplace bullying.
The weight of all these reports suggest that workplace bullying can fit any scenario at work, whether defined by a potential victim, a bystander, a researcher or commentator (Lewis, 2002).Workplace bullying is therefore a factor of the workplace, position of the bully and the bullied in the organisation.The types of workplace bullying experienced or reported occur as a result of the roles performed by the bullies or bullied within these organisations.This argument indicates a link between workplace bullying and organisational structure.This paper reports on the explored types of workplace bullying experienced by some employees in the ESO.

Method
The choice of questionnaire in this study was influenced by previous empirical studies that utilised the Negative Act Questionnaire (NAQ) developed by Einarsen and Raknes (1997).An adapted version of the Negative Act Questionnaire Revised (NAQ-R) and utilised in the Work Environment Survey (WES) by researchers at Griffith University was adopted in this research study.The questionnaire is an internationally accepted research instrument used by different researchers to explore various acts of bullying.The questionnaire consists of questions representing job satisfaction, various bullying acts, duration of the acts, causes of bullying, the effects of the bullying and actions taken to deal with their experiences.A combination of both online surveys and postal surveys were used to collect the data.The online survey made the process of collection less cumbersome in the sense that people responded faster and all the questions were answered, while the postal survey had a wider coverage, especially in some of the offices where the employees had no access to the on-line facilities.The NAQ-R used in this research was redesigned to suit the research context.The questions asked were based on the aim of the research, which was to explore the nature and the extent of workplace bullying.Questions related to the nature of bullying experienced were selected from the types of bullying that the employee experienced or witnessed.The whole organisation was targeted, and the questionnaires were distributed across all the 52 different locations of the organisation.The population of the organisation was 1,844 and a 25% response rate was achieved.The research was sensitive in nature and issues of confidentiality and anonymity were addressed in accordance with required ethical standards for research of this type.

Participants
The main data used for the analysis was based on the 452 returned questionnaires.The data revealed that 57.5 % of the respondents were male, and 41.5 % female.The respondents fell between the ages of 18 and 51 years, with 16.2 % under the age of 30, 65.5 % between the ages of 31 and 50, and 18.4 % 51 years and over.For the ethnicity category, 68 % of the respondents were White Welsh, 18.4 % White English, 1.5 % White Irish, 10.8 % other White background, and 0.4 % Black or Caribbean.Concerning sexual orientation, 90.7 % were heterosexual, 0.9 % lesbians, 2.8 % gay, and 0.7 % bisexual.With respect to disability, 36.9 % disabled, 25.7 % are not disabled, and 37.4 % preferred not to answer the question.For religion, 59.9 % were Christian, 0.4 % Hindu, 3.1 % Sikh, 6.4 % other, and 30.1 % reported that they had no religious beliefs.For length of service in the organisation, 17.9 % had been there for less than a year, 18.6 % between 1-5 years, 10.2 % 6-10 years, 20.8 % 11-15 years, 6.6 % 16-20 years, 11.7 % 21-25 years, and 14.2 % had worked there for 26 years and above.Concerning the work group to which the respondents belonged, 49.6 % were operational staff, 46.9 % non-operational staff, and 3.5 % indicated 'other'.The complete demographic information is shown below in Table I.

Measures
The questionnaire used in the study was labelled 'Work Environment Survey'.The word workplace bullying was not used in any part of the questionnaire; rather it was referred to as inappropriate behaviours, although, for the purposes of this research study, they connote the same meaning.Thus, in this paper, the terms will be used interchangeably.Data was collected on the 23 types of inappropriate behaviours experienced.Workplace bullying was labelled inappropriate behaviours in the workplace and defined as behaviours that are unreasonable, unacceptable or inappropriate.This structure and order followed an objective measurement of perceived exposure to specific bullying behaviours given by Rayner and Hoel (1997), Hoel et al., (2000), Salin (2003) and Einarsen (2006).The profiles of the employees and the different variables used to describe them are gender, age, age group, sexuality, disability, religion or beliefs, ethnic background, length of service, and group.This is again consistent and expected of the demographics of ESO, given that the majority of the workforce are white and less that 1 per cent of the workforce are non-white.However, given the type of statistical analysis conducted, this variable (ethnicity) was recreated by dividing the ethnic groups into two.That is, the majority ethnic group (white); and other minority groups were added together and recoded as (others).This was meant to reduce the bias of using data that is not evenly distributed.Moreover, the non parametric test conducted on these variables allows the use of data that are not normally distributed.
Having an inconsistent data set when compared with the demographics of ESO is one of the limitations of this study.However, given the rigour and transparency involved in the process of collecting information from the respondents, without ignoring the validity bias, the data collected can be said to be reliable.That is, the data collected has reflected a true pattern of interaction and relationship between the demographic composition, employee's exposure to bullying behaviours and the willingness to take part in the research.The distribution of these variables listed above showed that most of the variables deviated from the standard normal frequency distribution (see Figure 1) by having either a skewed distribution that is not symmetrical or a kurtosis distribution that is pointy or flat (Field, 2005).In the case of gender and nationality, the frequency distribution can be said to be positively skewed, while the sexuality and age are both negatively skewed.The disability, religious group and the length of service have a kurtosis distribution that is flat (platykurtic).While the operational group appears to be the only variable normally distributed.Given that most of the variables were not normally distributed, the non-parametric statistical test conducted on the variables was most appropriate given the characteristics of the data.

Types of Bullying Experienced: Factor Analysis
Factor analysis was carried out on the 23 negative behaviours highlighted in the questionnaire, to reduce the attribute space from a larger number of variables to a smaller number of factors.This analysis revealed similarities between some variables, thereby forming such similar variables into common factors.The sample size is 452, which is about 25 per cent of the population, and is adequate for this kind of analysis (Garson, 2007).Prior to conducting the Principal Component Analysis, two different test of suitability were conducted.First, a Pearson correlation coefficients test was conducted among all the identified negative behaviours in the NAQ-R.The correlation coefficients are all less than 0.9, the significance values of all the variables are greater than 0.05, and the determinant value of 1.28x10 -0.005 is greater than the necessary value of 1x10 -5 .The factor analysis showed that there is no multicollinearity between the variables, and all the variables correlate fairly well.Second, a Kaiser-Meyer-Olkin measure of sampling was conducted to check the pattern of correlation.A value close to 1 indicates that the patterns of correlations are relatively compact and will yield distinct and reliable factors, while values greater than 0.04 are acceptable (Field, 2005).Considering the percentage of variance accounted for by the factors, the Screeplot was examined.This plot, according to Field (2005), will show the relative importance of each factor extracted.This is done by plotting each factor in the factor analysis against the Eigenvalue.The screeplot (see Figure 2) revealed a break before the fourth and fifth component.Considering that only one item loaded in the fourth factor, I made a decision to use only three components for further investigation and dropped Factor 4 (Field, 2005).The Table 2 (see Appendix) shows the value to be 0.930, which falls, according to Field (2005), in the range of 'superb'; therefore, factor analysis is appropriate for these data.Bartlett's test of sphericity is aimed at testing the null hypothesis that the original correlation matrix is an identity matrix (Field, 2005).For this data, the level of significance is p<0.001, which indicates statistical significance; therefore, the R-matrix is not an identity matrix.It can therefore be said that there are relationships among the variables.These two tests indicate that factor analysis is appropriate for the data.
The aforementioned analysis is presented below in Table 2 and Table 3.The Loading of the Factors list the four components within the data.The eigenvalues associated with each linear component (factor) before extraction identified that Factor 1 explains 62.2 per cent of the total variance, Factor 2 explains 7.6 per cent of the total variance, Factor 3 explains 6.9 per cent of the total variance, and Factor 4 explains 4.4 per cent of the total variance (see Appendix 8).The SPSS software was therefore set to extract four factors.The first factor is very high (62.2 per cent), which is expected of any factor analysis (Field, 2005).The rotated sum of squared loading optimises the effect of the factor structures; that is, the importance of each of the four factors is equalized.The rotated Factor 1 now accounts for 28.7 per cent of total variance, Factor 2 accounts for 26.1 per cent of the total variance, Factor 3 accounts for 17.8 per cent of the total variance, and Factor 4 accounts for 8.7 per cent of the total variance (see Table 4).The rotated component matrix in Table 3) displays all loadings greater than 0.4.The suppression of the loading is aimed at making the interpretation easier (Field, 2005).All the values in this table are greater than 0.4, which implies that all the variables included in the analysis are very important and they all contribute substantively to the factors.The internal consistency of each component was assessed with the Cronbach alpha.A value lower than 0.7 according to Field ( 2005) is considered as an unreliable scale, however, he gave an exemption on psychological construct, with values less than 0.7 as realistic because of the diversity of the construct being measured.In line with this analysis, alphas of 0.93, 0.91 and 0.88 for each of the three components were found, which suggest reliable consistency of the components.
The content of the questions that load into one factor led to the identification of common themes, which are given new labels.Looking at Factor 1, the four issues with the highest values or loadings are as follows: Being subjected to inappropriate materials in the workplace, for example, posters (0.843) Having insulting or offensive remarks made about your person, your attitudes or your private life (0.752) Threats of violence or physical abuse (0.746) Being the subject of excessive teasing and sarcasm (0.739)All these issues seem to relate to cynical and personal acts.Before these loadings were labelled, the other variables that fell under Factor 1 and were correlated are as follows: intimidating behaviour such as finger pointing, hints or signals from others that you should quit your job, repeated reminders of your errors or mistakes, and practical jokes carried out by people you do not get along with.Based on these results, this factor was labelled personal bullying.
Looking at the second factor, the four issues that are loaded highly on Factor 2 are as follows: Having key areas of responsibilities removed or replaced with unpleasant tasks (0.787) Being ordered to work below your level of competence (0.779) Excessive monitoring of your work (.584) Pressure not to claim something which by right you are entitled to, for example, sick leave, holiday entitlement, travel expenses (0.563) These issues seem to relate directly to the work situations, and it can be seen that they are behaviours out of the control of certain individuals.Some other variables under Factor 2 that were highly correlated are as follows: being exposed to an unmanageable workload, excessive monitoring of your work, and persistent criticism of your work.This factor was therefore labelled administrative bullying.
In the third factor, the three issues that load highly are as follows: Someone withholding information (0.757) Having your opinions and views ignored (0.602) Being shouted at or being the target of spontaneous anger or rage (0.509) Similarly, the other variables that fell under Factor 3 and were correlated are as follows: being ignored or facing hostile reactions when you are approached, persistent criticism, being ignored and excluded from activities.These issues seem to relate to scapegoating, isolation and being singled out by the group.This factor was therefore labelled social exclusion.

Hypothesis Testing: The Wilcoxon Mann-Whitney Test
In the previous analysis, three main types of workplace bullying were identified: personal bullying, administrative bullying and social exclusion.A Wilcoxon Mann-Whitney Test was carried out on these three types of bullying and some of the demographic factors such as gender, sexual orientation, age, rank, length of service in the organisation, ethnicity and disability.Although demographic characteristics can be used to explain why certain employees are more likely to become bullies or victims of bullying, there is still a gap in the literature concerning whether there are significant differences in the types of bullying to which employees are exposed based on demographic differences.The rationale for addressing this issue allows the argument concerning whether certain groups of people are more or less likely to experience a type of bullying than are others (Adewumi, Sheehan & Lewis, 2008).Given this conception, hypotheses were proposed and tested to determine the differences in the bullying experienced as a factor of demographic differences.All the factors mentioned above are expected to play a significant role in the kind of bullying behaviours to which employees are exposed.The hypotheses developed in this study were based on the reviewed literature on workplace bullying.The results are presented in accordance to the hypotheses stated in the previous chapter.
Revisiting the hypotheses, the relationship between personal bullying and all the demographic factors are explored.As a result, the following hypotheses are proposed: Hypothesis H1: There are no statistically significant differences in the exposure of employees to personal bullying.
The main hypothesis is further divided into sub-hypothesis, and it is expected that there should be significant differences in the exposure of employees to personal bullying.Hypothesis H1a: Gender plays no role in the exposure of employees to personal bullying Hypothesis H1b: Ethnicity plays no role in the exposure of employees to personal bullying Hypothesis H1c: Age plays no role in the exposure of employees to personal bullying Hypothesis H1d: Sexuality plays no role in the exposure of employees to personal bullying Hypothesis H1e: Disability plays no role in the exposure of employees to personal bullying Hypothesis H1f: Length of service in the organisation plays no role in the exposure of employees to personal bullying Hypothesis H1g: Occupational group within the organisation plays no role in the exposure of employees to personal bullying The Table 5: Personal Bulling: Test of Significance below presents a test of the differences among the different groups in regard to their exposure to personal bullying.For personal bullying, the results show that there are statistically significant differences among the various groups.Thus, for the main Hypothesis 1, the results show that there are significant differences in the exposure of employees to personal bullying.Hence, the null hypothesis that there are no significant differences is rejected and the alternative hypothesis is accepted.For the sub-hypotheses, the results are collectively presented.For gender, the test statistics show that the p value (0.000) is less than 0.01 (99% confidence predetermined level of significance) and the mean rank for male (231.50) is greater than that for female (190.28).These findings imply that men are more exposed to personal bullying when compared to women in the organisation.For the two occupational groups within the organisation, the p value (0.000) is less than 0.01, and the mean rank for operational staff (226.9) is greater than that for support staff (192.54).Hence, the groups with a higher mean rank value are more exposed to personal bullying when compared to the others.Variables such as ethnicity and length of service are significant at the 0.05 predetermined level of significance (95% confidence predetermined level of significance).
The remaining variables, such as sexual orientation (gay and heterosexual), age (age groups under 30 and 51 years or over) and disability (disabled and non-disabled employees) all show non-significant differences in their exposure to personal bullying.Thus, the hypotheses that there are no significant differences with respect to gender (H1a), ethnicity (H1b), length of service (H1f) and occupational group (H1g) are rejected and the alternative hypotheses are accepted; while the hypotheses that there are no significant differences concerning age (H1c) and disability (H1e) are accepted and the alternative hypotheses rejected.

Similar to the first hypothesis, the second hypothesis is stated as follows:
Hypothesis H2: There are no statistically significant differences in the exposure of employees to administrative bullying.This is further divided into sub-hypothesis; however, it is expected that there should be significant differences in the exposure of employees to administrative bullying.Thus, the sub-hypotheses are listed as follows: Hypothesis H2a: Gender plays no role in the exposure of employees to administrative bullying Hypothesis H2b: Ethnicity plays no role in the exposure of employees to administrative bullying Hypothesis H2c: Age plays no role in the exposure of employees to administrative bullying.Hypothesis H2d: Sexuality plays no role in the exposure of employees to administrative bullying Hypothesis H2e: Disability plays no role in the exposure of employees to administrative bullying Hypothesis H2f: Length of service in the organisation plays no role in the exposure of employees to administrative bullying

Hypothesis H2g: Occupational group within the organisation plays no role in the exposure of employees to administrative bullying
The results of this analysis are illustrated in Table 6 below.Looking at the variable administrative bullying, the results show that there are statistically significant differences among the various groups, as proposed.Thus, for the main Hypothesis 2, the results show that there are significant differences in the exposure of employees to personal bullying.Hence, the null hypothesis that there is no significant difference is rejected and the alternative hypothesis is accepted.However, for the sub-hypotheses under administrative bullying, gender has a p value = 0.000 and the mean rank for male (240.76) is greater than that for female (179.33).Sexual orientation has a p value = 0.002 and the mean rank for gay (120.11) is less than that for heterosexual (203.42).For age, the p value = 0.000 and the mean rank for employees under 30 years (62.11) is less than that for those 51 years or over (91.01).Disability has a p value = 0.004 and the mean rank for non-disabled (144.32) is greater than that for disabled (112.48).For ethnicity, the p value =0.000 and the mean rank for white (164.32) is less than that for other background (232.40).Thus, the hypotheses that there are no significant differences with respect to gender (H2a), ethnicity (H2b), age (H2c), sexuality (H2d), disability (H2e), length of service (H2f) and occupational group (H1g) are rejected and the alternative hypotheses are accepted.These results imply that men, those aged 51 years or over, heterosexuals, non-disabled people, and those with an ethnic background described as 'other' are more exposed to administrative bullying when compared to others.The higher the value of the mean rank, the more exposed the groups are to administrative bullying.Other variables such as occupational group and length of service are significant at the 0.05 predetermined level of significance.

Hypothesis H3: There are no statistically significant differences in the exposure of employees to social exclusion
The hypothesis is further divided into sub-hypotheses, and it is expected that there should be significant differences in the exposure of employees to social exclusion.The sub-hypotheses proposed are as follows: Hypothesis H3a: Gender plays no role in the exposure of employees to social exclusion Hypothesis H3b: Ethnicity plays no role in the exposure of employees to social exclusion Hypothesis H3c: Age plays no role in the exposure of employees to social exclusion Hypothesis H3d: Sexuality plays no role in the exposure of employees to social exclusion Hypothesis H3e: Disability plays no role in the exposure of employees to social exclusion Hypothesis H3f: Length of service in the organisation plays no role in the exposure of employees to social exclusion Hypothesis H3g: Occupational group within the organisation plays no role in the exposure of employees to social exclusion The results of the analyses are illustrated in Table 7 below.The results show that there are statistically significant differences among the various groups to social exclusion, as proposed.Thus, for the main Hypothesis 3, the results show that there are significant differences in the exposure of employees to social exclusion.Hence, the null hypothesis, which states that there are no significant differences, is rejected and the alternative hypothesis is accepted.However, for the sub-hypotheses under social exclusion, gender has a p value = 0.000 and the mean rank for male (237.07) is greater than that for female (184.10).Sexual orientation has a p value = 0.002 and the mean rank for gay (106.25) is less than that for heterosexual (203.92).Age has a p value = 0.000 and the mean rank for employees aged under 30 years (62.41) is less than that for those aged 51 years or over (90.75).Also of interest is disability with a p value = 0.000, with the Mean Rank for non-disabled (147.77)being greater than that for disabled (106.64).Ethnicity has a p value = 0.000 and the mean rank for white (167.27) is less than that for other background (212.23).Occupational group has a p value = 0.002 and the mean rank for operational staff (197.08) is less than that for support staff (228.13).Thus, the hypotheses that there are no significant differences with respect to gender (H3a), ethnicity (H3b), age (H3c), sexuality (H3d), disability (H3e), length of service (H3f) and occupational group (H3g) are rejected and the alternative hypotheses are accepted.However, these results imply that heterosexual men, those aged 51 years or over, non-disabled people, those with an ethnic background described as 'other', and those who are support staff are more socially excluded than the other groups of people.
Hence, from the above analyses, it can be concluded that some of the demographic factors mentioned above play significant roles in the exposure of employees to bullying behaviours (Adewumi et al., 2008).Although most of the literature reviewed has looked at the effect of these factors on the likelihood of exposure to bullying (see Eriksen & Einarsen, 2004;Hoel & Cooper, 2000;Leymann, 1996;Salin, 2003;WBTI, 2003), the present study has significantly contributed to existing knowledge on workplace bullying by revealing that the type of bullying behaviours to which employees are exposed is a factor of demographic factors and characteristics, as explained above.

Discussion
The major aim of this paper was to explore the types of workplace bullying behaviours experienced by some staff in an emergency services organisation in the UK.Three main types of behaviours that could be seen to constitute workplace bullying were identified.They are personal bullying, administrative bullying, social exclusion and humiliation.This classification is closer to those identified by Vartia (2001).The findings from the factor analysis indicate that for those components that are classified as personal bullying, the percentage of employees who reported having been exposed to threats of violence or physical abuse is low when compared to forms of negative acts (Adewumi, Sheehan & Lewis, 2007).One explanation is that in the United Kingdom, there is legislation related to assaults and physical attacks, where people can take the necessary action when attacked physically in the workplace, unlike other bullying acts that are not addressed by discrete legislation.Other studies have confirmed similar low rates of physical attack or abuse (Hoel & Cooper, 2000;Rayner & McIvor, 2006).In the case of verbal abuse, employees in this study have reported insulting and offensive remarks being made about them in the workplace.These findings about verbal abuse are supported by other empirical findings that suggest that verbal abuse is one of the most reported types of workplace bullying (see Hoel et al., 2001;Mikkelsen & Einarsen, 2001;Rayner & Cooper, 2003).
In the case of administrative bullying, this type of workplace bullying could occur as a result of various factors, such as the tasks performed in the organisation and different levels and positions occupied by the employee(s) within the organisation.For instance, in the study carried out by Hoel and Cooper (2000), they reported that the roles performed by supervisors or managers and subordinates can be a factor of the type of bullying they experience and a similar report was given by Lewis and Gunn (2007).Hence, it is important to note that when stating bullying is administrative, this indicates it is work related and the management of ESO could be playing a vital role in the work relationship (Adewumi et al., 2007).In an organisation where there is a great degree of control and the management of employees centres on issues such as stability, rules, division of labour and hierarchical structures, there will be cases of employees subjected to such behaviours (see Thompson & McHugh, 2002).Thus, the finding here that employees have their opinions ignored, are ordered to work below their level of competence and have key areas of responsibility removed or replaced with unpleasant tasks, is simply pointing to the fact that workplace bullying can be induced by the management or the organisation (Adewumi et al., 2008).
For social exclusion, the results show that respondents reported having had information withheld from them and their opinions and views ignored.Researchers such as Leymann (1996), Salin (2006) and Vartia (2003) have all reported social exclusion as one of the types of bullying.The social exclusion reported in this case can be attributed to many factors, such as the demographic composition (gender, ethnicity, sexual orientation) and the operational division of the organisation into uniformed and non-uniformed staff.That is, group composition may be considered to influence the type of bullying experienced (Adewumi et al., 2007a).
Given that being a member of a dominant group in the work environment can provide an identity and self-categorisation, especially in a situation determined by social identification and consequent behaviour (Capozza & Brown, 2000), those who are not categorised as members of the group can find themselves singled out or socially excluded from activities.Keeping people away or uninformed about basic issues can be perceived as being singled out or left out of activities and tasks.These acts, according to Vartia (2003), move or change the bullying process, which normally begins with indirect negative behaviours, for example, gossip and malicious comments, which in most cases are difficult to counteract into a more direct form, such as exclusion.
Although this paper is not exploring the types of bullying experienced by employees in different positions or ranks within the organisation, it is important to note that, when we say bullying is administrative, the organisation and the management of the organisation are playing a vital role in that relationship.The findings here reported have shown that employees are having their opinions ignored, suggesting that in this organisation, you have to be in a certain position or rank within the organisation before your ideas and contributions are seen as important.
These findings were brought to the notice of the management of the organisation, who at the initial start of this research were in denial of employees being bullied at work.To buttress their point, there has been lack of evidence to support the report of the prevalence of workplace bullying in the organisation.However, previous empirical studies such Archer (1999) and Thompson and McHugh (2002) have reported that in an organisation where there exists a command and control structure in order to facilitate stability, predictability of work force behaviour, rules, division of labour, lines of work, and job identity, there will be cases of employees subjected to such behaviours.Given that workplace bullying is going in ESO, some of the preliminary investigations showed that employees are willing to work till retirement because ESO is a public sector organisation where good conditions of employment are adhered to.

Practical Implication
It may be concluded from this research that bullying is part of the culture of this organisation, and that may be why it is perceived to be accepted as a norm and is continuing.As for the types of bullying experienced, especially those within the control of individuals (personal bullying and social exclusion), individual counselling might be appropriate for the victims and assistance from specialist counsellors on how to deal with their experiences might also be very useful.Counselling may not only help people to get back to their normal lives, but also may help them to deal with any future workplace bullying.
For the perpetrators, counselling and assistance is also required.Continuous counselling and training ought to be provided, because some bullies do not perceive themselves as bullies.In many situations the alleged bullies tend to justify their actions and blame it on the victims or other external factors such as the need to be firm and in control, or even organisational demands (Sheehan, 2006).Bullies in this situation may not be aware of what they are doing or how their behaviours may affect others.
The intervention against such behaviours should be aimed at the victim(s), the perpetrator(s) and the management of the organisation.Putting in place intervention mechanisms, such as counselling, training on equality and diversity, and emphasising the effects of bullying on the employees and the organisation, may help to create a continuous awareness of this social phenomena.Furthermore, a zero tolerance policy ought to be introduced with appropriate sanctions and outcomes, including how to manage vexatious complaints, so that perpetrators will not get away with bullying.In a situation where employees are trained on issues related to employees' well-being, equality and diversity, harassment and bullying, there should be constant refresher or reminder courses at close intervals, so as to ensure that all the workforce understands and endorses appropriate ways of working together.
Finally, since this organisation is characterised by a very rigid structure, the managers need to be more aware of the effects of bullying, for instance, the effect on the individual or the cost effect on the organisation.Such awareness could be a bedrock on which managing relationships at work could be built.The management of the organisation ought to have in place clear policies on workplace bullying.A clear and precise policy on bullying will ensure that the victims are protected and the perpetrators are easily identified.Proper signals should be sent across the organisation, so that the employees will have a clearer picture of the roles played by management in addressing the problem.

Limitations and Assumptions of the Study
Although the present study has shed some light into workplace place bullying in an emergency service organisation in the UK, there are several limitations that must be kept in mind.First, the study from which this paper was drawn was not a highly controlled empirical study because it was designed to be an exploratory study, and opened to any discovery of new insight.However, this limitation was minimised because most of the analyses were based on prior knowledge of workplace bullying occurring in an organisation characterised by power, ranks and hierarchy (See Archer, 1999).Also previous work done by researchers such has Lewis (1999) in the UK and Vartia (2003) in the Scandinavian country have reported workplace bullying to be high in public sector organisations characterised by formal power differences such as grades found in a highly structured organisation.Second, the study is based on one particular organisation; hence the findings might not be applicable to other organisations.Third, although the results indicated the types of bullying, the analysis did not show the relationship between the bullying experienced and the demographic characteristics of the respondents.A future paper will explore the relationship, so as give a clearer picture of the experiences of bullying.
Since the study was based on strict anonymity, for those respondents that have been bullied and are still experiencing the behaviours, it will be difficult to offer them one to one feedback, mediation, or other intervention measures.Unless they identify themselves, the organisation might be restricted in being able to give direct support to those people.The victims may only benefit from the general interventions given to all staff, and those interventions might not serve an individual purpose.

Table 3 .
Loading of the Bullying Factors Pressure not to claim something which by right you are entitled to(e.g.sick leave, holiday entitlement, travel expenses)

Table 5 .
Personal Bulling Test of Significance

Result of Wilcoxon Mann -Whitney Test of Significance Significant Differences Significant Differences Significant Differences
Difference is significant at 0.05 level (2 -tailed) ** Difference is significant at 0.01 level(2 -tailed) *