Description
Workplace bullying has been increasingly acknowledged as a major concern to British employers and their workers, with implications for individuals as well as organisations.
Destructive Interpersonal Conflict
in the Workplace:
The Effectiveness of Management
Interventions
Funded by the
British Occupational Health Research Foundation (BOHRF)
Helge Hoel & Sabir I. Giga
With contributions from Brian Faragher
Manchester Business School
The University of Manchester
February, 2006.
©Helge Hoel & Sabir I. Giga, 2006.
2
Contents
Acknowledgements.................................................................................................... 4
Executive Summary................................................................................................... 5
1.1 Introduction.......................................................................................................... 7
1.2 Background.......................................................................................................... 8
1.3 Key findings from previous intervention research................................................ 9
2. Research method................................................................................................. 10
2.1 Focus groups.................................................................................................. 11
2.2 Questionnaire survey..................................................................................... 11
2.3 Interventions................................................................................................... 12
2.4 Research design............................................................................................ 12
2.5 Local steering committees.............................................................................. 13
3. Pre-intervention focus groups.............................................................................. 14
4. Development of the Bullying Risk Assessment Tool (BRAT)............................... 15
4.1 Factor structure.............................................................................................. 17
4.1.1 ‘Organisational fairness’........................................................................... 17
4.1.2 ‘Team conflict’ .......................................................................................... 17
4.1.3 ‘Role conflict’............................................................................................ 18
4.1.4 ‘Workload’ ................................................................................................ 18
4.1.5 ‘Leadership’ ............................................................................................. 19
4.2 Factor reliability.............................................................................................. 19
4.3 Validation of the BRAT................................................................................... 20
4.4 Application of the BRAT ................................................................................. 21
5. Identification and development of interventions................................................... 21
5.1 Policy communication..................................................................................... 22
5.2 Stress management training programme....................................................... 22
5.3 Negative behaviour awareness training programme...................................... 23
5.4 Implementation of interventions...................................................................... 24
6. Questionnaire survey participants and sample background................................. 25
6.1 Sample - study participants............................................................................ 25
6.2 Sample background....................................................................................... 26
6.2.1 Gender..................................................................................................... 26
6.2.2 Age .......................................................................................................... 27
6.2.3 Ethnicity ................................................................................................... 28
6.2.4 Contracted hours of work......................................................................... 28
6.2.5 Employment contract ............................................................................... 29
6.2.6 Job type ................................................................................................... 29
6.2.7 Supervisory responsibilities ..................................................................... 29
6.2.8 Length of time with organisation .............................................................. 30
7. Objective data...................................................................................................... 30
8. Experiences of bullying in the workplace: baseline measures ............................. 31
8.1 Experience of bullying.................................................................................... 31
8.2 Witnesses of bullying...................................................................................... 33
8.3 Past bullying................................................................................................... 33
8.4 Targets of bullying.......................................................................................... 34
8.4.1 Gender..................................................................................................... 34
8.4.2 Age .......................................................................................................... 34
8.4.3 Full-time versus part-time ........................................................................ 35
8.4.4 Ethnicity ................................................................................................... 35
3
8.5 Organisational level of target.......................................................................... 36
8.6 Perpetrators of bullying .................................................................................. 36
8.7 How many were bullied? ................................................................................ 37
8.8 Duration of the bullying experience................................................................ 37
8.9 Experiences of negative behaviours............................................................... 37
8.10 Outcomes of bullying.................................................................................... 38
9. Experiences of bullying in the workplace pre and post intervention
comparisons .................................................................................................... 39
9.1 Experiences of bullying.................................................................................. 39
9.2 Perpetrators.................................................................................................... 40
9.3 How many were bullied? ................................................................................ 41
9.4 Psychological contract.................................................................................... 42
10. Efficacy of intervention: comparing baseline and post-intervention
measures for different interventions across organisations............................... 43
10.1 Overall results .............................................................................................. 43
10.2 Bullying......................................................................................................... 44
10.3 Witnessing bullying....................................................................................... 45
10.4 Negative behaviours..................................................................................... 46
10.5 The psychological contract........................................................................... 47
10.6 Mental health................................................................................................ 49
10.7 Organisational outcomes, absenteeism and intention to leave..................... 50
10.8 Satisfaction with the atmosphere at work..................................................... 51
10.9 The effectiveness of intervention: assessing ‘objective’ measures............... 53
10.10 Intervention and levels of absenteeism...................................................... 53
10.11 Other ‘objective’ measures......................................................................... 54
10.12 Total assessment of effectiveness of interventions .................................... 54
11. Intervention feedback from the trainer and participants ..................................... 56
11.1 The trainer’s perspective.............................................................................. 56
11.2 Feedback from participants.......................................................................... 57
12. Post Intervention focus groups........................................................................... 58
13. Discussion ......................................................................................................... 63
14. Conclusions ....................................................................................................... 68
15. References......................................................................................................... 70
16. Appendices........................................................................................................ 74
Appendix A: Advisory Board Membership............................................................ 74
Appendix B: The Bullying Risk Assessment Tool (BRAT).................................... 75
Appendix C: Validation of the BRAT (results of regression analysis)................... 77
Appendix D: Policy guidelines.............................................................................. 78
Appendix E: Part A Tool: Bullying Risk Assessment Tool .................................... 80
Appendix F: Demographic variations in the experience of bullying to bullying..... 81
Appendix G: Pre- and post-intervention data comparisons .................................. 84
4
Acknowledgements
We would like to thank the British Occupational Health Research Foundation
(BOHRF) for funding this research and in particular, Brian Kazer, the Chief
Executive of BOHRF for his unremitting support throughout the duration of
this project. Our unreserved gratitude to members of our Advisory Board
(Appendix A), not least the Chair – Baroness Gibson of Market Rasen, for
their guidance and constructive critique during various stages of the research.
We would also like to acknowledge the support of the five organisations who
took part in this study for not only agreeing to participate but also for
demonstrating enduring commitment, and to Rob Allan and the West
Hertfordshire NHS Trust for allowing us to pilot our interventions within their
organisation. Our appreciation also to Professor Cary Cooper for his support
in initiating this study as well as his continued interest in our work. Finally, we
would like to thank our colleague Dr Brian Faragher for his invaluable advice
on statistical analysis and our partners Neil and Tisha for their everlasting
patience.
5
Executive Summary
Workplace bullying has been increasingly acknowledged as a major concern
to British employers and their workers, with implications for individuals as well
as organisations. Whilst most studies to date have focused on the prevalence
of the phenomena and its consequences, very few have looked at remedial
action. To address this imbalance, an intervention study was carried out in
five British public sector organisations.
Sponsored by the British Occupational Health Research Foundation (BOHRF)
the aim of the study was two-fold: 1) to devise and test the appropriateness as
well as effectiveness of a risk assessment tool, and 2) to develop, implement
and evaluate three different bullying intervention programmes. These
programmes focussed on training in three different areas: policy
communication, stress management and negative behaviour awareness.
With regard to study method, a randomised control design was deemed
necessary to enable the researchers to make any degree of inference with
regard to causal relationships, with the same interventions carried out in
various combinations in all five organisations. Pre and post intervention data
were obtained by means of a questionnaire comprising of a variety of
instruments to measure negative behaviour and experiences and
consequences of bullying. This was supplemented with data from post-
intervention focus groups and information from the trainer as well as training
participants. In line with the study’s aims, the questionnaire survey also
contained an instrument to measure potential risk-factors of bullying – the
Bullying Risk Assessment Tool (BRAT). Informed by focus group findings as
well as expert opinion and a literature review, the refined version of the BRAT
has a factor structure made up of five factors: organisational fairness, team-
conflict, role-conflict, workload and leadership. A validation study confirmed
the validity of the instrument with all five factors emerging as predictors of
negative behaviour and self-labelled bullying.
Pre-intervention base-line measures were based on a sample of 1,041
respondents, a response-rate of 41.5%. The pre-intervention questionnaire
study confirmed that bullying is a problem in public sector organisations in the
UK with 13.6% of respondents stating that they have been bullied within the
last six months, compared with a national average of 10.6% obtained from a
previous BOHRF sponsored nationwide study (Hoel & Cooper, 2000).The
level of bullying in the current study varied between organisations, with the
lowest at 10.8% and highest at 16.6%. An analysis of the data highlights a
strong association between bullying and mental health and intention to quit,
thus confirming that bullying appears to have individual as well as
organisational implications. A total of 61% of perpetrators were reported to be
managers or supervisors whilst colleagues were considered to be the culprits
in 42% of incidents.
The main stress management and negative behaviour awareness
interventions were delivered by a professional trainer as three-hour training
6
sessions, whilst the policy communication was carried out over a thirty-minute
session. The training programme was implemented in all five organisations
over a six month period and involved approximately 150 participants in total.
Feedback contained in 193 training evaluation forms (participants on the full-
day training programme were required to evaluate the morning and afternoon
sessions separately), suggests that the training was well received, with a
substantial number of participants considering the sessions very interesting as
well as relevant. With regard to the trainer’s own end-of-training session
comments, it was emphasised that sufficient time was needed in order to
establish a climate in which constructive group dynamics and learning could
take place.
Post-intervention measures were obtained approximately six months after the
training had taken place by means of distributing the same questionnaire a
second time within the same units in all five participating organisations. This
made it possible to compare post and pre-intervention data for a large number
of variables relevant to the issue of bullying and negative behaviour at work.
Important improvement in the desired direction as measured by a variety of
variables did occur for 45% of the experiment groups. For three of the
experiment groups scores on all relevant variables measured were in the
desired direction. In two of these cases the experiment group had received a
combination of all three training programmes or workshops whilst the third
group had received the ‘negative awareness training’ in addition to policy
communication. The fact that it is difficult to make any firm conclusions with
regard to efficacy of a specific intervention is not surprising and is in line with
much organisational research. Threat of redundancy and in some cases a
strong unwillingness to engage with the training might also help explain why
the desired development did not take place in all experiment groups. Other
factors that may have possibly influenced the results include: increased
awareness and greater expectations which in turn lead to dissatisfaction
resulting from the training process; the right people, in this case managers
and supervisors, not taking part in the training in sufficient numbers; overall
‘critical mass’ not being achieved due to too few people being trained in order
to have a significant impact upon behaviour; and, finally, the time between
interventions and post-intervention measurement might have been too short
for significant effects to have occurred.
In conclusion, this research has resulted in the successful completion of the
first academic anti-bullying intervention study, comparing the effectiveness of
interventions across different organisational contexts and involving the
implementation of a complex design in order to apply scientific rigor. A risk
assessment tool focussing on issues relating to negative behaviour and
bullying has been developed and its properties validated. Although the study
was unable to establish beyond doubt the efficacy of a particular intervention,
there is evidence to suggest that theoretically sound, well planned and aptly
delivered interventions can make a difference, particularly when sufficient time
is allocated and the proportion of staff being trained is significant enough to
have an impact upon behaviour.
7
1.1 Introduction
This report presents the results of the very first academic study evaluating the
effectiveness of a variety of interventions to tackle workplace bullying and
highlights the opportunities organisations have in preventing and managing
this problem. Following on from the previous collaboration between the
University of Manchester Institute of Science and Technology (UMIST) and
the British Occupational Health Research Foundation (BOHRF) which
established the prevalence of workplace bullying (the first such study of its
kind in Great Britain), and having identified a number of sources associated
with negative behaviour in the workplace, the overall aim of this project was to
develop, implement and evaluate the effectiveness of organisational
interventions to prevent and manage destructive interpersonal conflict.
Based on the successful application of risk assessment to physical hazards in
the workplace, recent developments have seen attempts to apply such
techniques to the area of psychosocial hazards. In addition to identifying
effective intervention strategies in the present study, we also aimed to take
advantage of these theoretical gains to develop a risk assessment tool which
could assist managers and their organisations in managing and controlling
antecedents of destructive behaviour and bullying in the workplace.
We envisage that our findings will help inform academic and organisational
practice in the development of effective intervention and risk management
strategies to deal with interpersonal conflict in the workplace and as such
make an important contribution to the development of this relatively new field
of research and practice.
Structure of the report:
After providing a brief background to the study, the methodology is outlined in
section 2. The next section reports on the pre-intervention focus groups
carried out in order to inform the interventions as well assist in the
development of a risk assessment tool. Section 4 focuses on the development
and validation of the Bullying Risk Assessment Tool (BRAT) and section 5
examines the development and implementation of various interventions
identified to combat bullying and destructive behaviour. The study sample and
sampling procedure is then outlined in section 6. The findings of the two
surveys, baseline and post-intervention data, with measurements of a number
of variables associated with bullying and destructive behaviour are
investigated in sections 8 and 9. With the aim of establishing potential efficacy
of particular interventions or combination of interventions, a number of
analyses of the data were carried out. The results of these analyses are
critically examined in section 10. This is followed by a presentation of the
qualitative data based on feedback from study participants and other
observers, including comments from the trainer. Finally, in the closing
sections we discuss the results and highlight the main learning points for
future intervention studies before reviewing our key findings, and delivering
our closing opinions and conclusion.
8
1.2 Background
A number of studies, in the UK and abroad, have recently identified
destructive conflict and bullying at work as an occupational hazard of
significant magnitude (Zapf et al., 2003). However, it was not until the first
BOHRF funded study (Hoel & Cooper, 2000) that it was possible to establish
the prevalence of the problem in the UK with a high degree of certainty taking
into account the size of the sample and the research methodology applied.
The data from this survey suggests that approximately 10% of the UK
workforce consider themselves at any one time to be bullied. For a majority of
these targets of bullying, the experience is reported to be occasional rather
than regular. If the time frame is changed to include experience within the last
five years, approximately 25% of respondents report themselves as having
been bullied. Furthermore, when observers or witnesses of bullying, who are
not themselves targeted directly, are included, nearly half the respondents
report having had some first-hand experience of the problem.
The nationwide study sponsored by BOHRF also confirms previous findings
which suggest that being exposed to destructive conflict and bullying may
have implications for individuals (see Einarsen et al., 2003), not only with
regard to their mental and physical health, but also with respect to their
organisational commitment and satisfaction. Industry-specific analysis also
suggests that the same behaviours may be present (though to a varying
degree) but are interpreted in various ways and have a different impact within
different organisational settings (Hoel, et al., 2004). This result is in line with
recent findings in stress research, which emphasise that the relationship
between work-related stressors and strain (i.e. ill health) should be
understood in a local or situation-specific context (Giga et al., 2003; Sparks &
Cooper, 1999).
With regard to organisational implications, the findings from the nationwide
study (and studies carried out by Charlotte Rayner for Unison, e.g. Rayner,
1999) suggest a relationship between bullying and intention to leave the
organisation, with targets of bullying substantially more likely than non-targets
to leave. Furthermore, for the first time a large-scale study has identified a
similar, albeit weaker, association between exposure to bullying, on the one
hand, and absenteeism and productivity, on the other, suggesting that
employees who are targets of bullying may on average have seven days more
sick-leave than those who were neither bullied nor had witnessed bullying.
Moreover, the negative effects of bullying appear to extend beyond those who
report being bullied at present to include those who were bullied in the past as
well as those who have witnessed bullying taking place within the same time-
frame, but who have not themselves been directly targeted (Hoel, Faragher &
Cooper, 2003). It could, therefore be argued that it should be in the self-
interest of organisations to take this issue seriously and attempt to manage it
effectively.
Whilst the antecedents of bullying will vary between organisations, most
organisations at the beginning of the new millennium are still in the midst of or
struggling with the effects of significant change processes. External pressures
9
have forced organisations to restructure, downsize and become leaner (Littler
et al, 1994). This is true for the private as well as public sector. As a result,
employees at all levels of the organisation, in the private as well as in the
public sector find themselves in a position of increasing workloads, often in a
climate of uncertainty with regard to their current and future employment
(Stewart & Swaffield, 1997). In this situation managers are frequently required
to put in long and even excessive working hours whilst accountability and
responsibility for the human resources function is often devolved to the level
of the line manager (J ames, 1993). Faced with this situation, it is not
surprising to find that managers allegedly frequently make use of authoritarian
and even abusive behaviours in order to carry out their work (Sheenan, 1999).
In light of such pressures it might come as no surprise that a common finding
from research into bullying in the UK, that in three out of four incidents the
alleged perpetrator appears to be someone who has supervisory or
managerial responsibilities (Unison, 1997; Hoel, Faragher & Cooper, 2001).
Most studies of workplace bullying have so far focused on establishing the
prevalence of the phenomena and its consequences By contrast, despite
substantial effort by practitioners in addressing the issues in the workplace
(e.g. Rayner, Hoel & Cooper, 2002; Hubert, 2003) little evidence has emerged
with regard to effective interventions. Furthermore, references to
organisational interventions within the emerging bullying literature tend to rely
on propositions inferred from related areas such as workplace violence,
sexual harassment and stress (Hoel et al., 1999). According to the literature
on interventions it is noteworthy that the evidence with regard to efficacy, in
particular with respect to psychosocial issues, is sparse, with a number of
observers highlighting the need for further attention to such issues (e.g.
Kompier et al., 1998; Cox et al., 2000; Murphy & Sauter, 2003). Thus, to our
knowledge there are no studies that have systematically attempted to assess
the efficacy of various forms of interventions aimed at reducing and
minimizing negative behaviour and bullying at work. Moreover, the fact that
incidence studies of workplace bullying in the UK consistently suggest that the
majority of perpetrators are to be found within supervisory and managerial
ranks suggests that behaviour of managers should be a focal point of any
intervention study. Therefore, such an approach would also seem to be the
most appropriate one from a cost-benefit perspective where resources are
limited.
1.3 Key findings from previous intervention research
By means of a review of the intervention literature (e.g. Cooper & Cartwright,
1997; Kompier et al., 1998; Cox et al., 2000; Giga et al., 2003; Murphy &
Sauter, 2003) we identified, and observed as far as possible in the present
study, a number of issues which were considered crucial in order to achieve a
successful outcome:
Most interventions studies focus on either the individual (predominantly)
or the organisation. Very few studies combine measures aimed at both
the individual and the organisation.
10
For any intervention to become successful it must be based on sound
theory. Moreover, this understanding cannot be the sole domain of
researchers but must be shared with the organisation and its
management.
Without commitment and support from senior echelons within
management, including provision of necessary resources and a clear
understanding of the theoretical underpinning of the interventions any
initiatives are likely to fail. The fact that the presence of bullying is
considered as ultimately resting on managerial consent also further
strengthens the case for active management involvement.
Current research methods are insufficient and progress in this area is
necessary in order to establish efficacy of interventions. In particular, it
is crucial to make use of more objective measures (data which can be
validated by others).
Any intervention needs to be contextualised and tailored to the
requirements of the organisations by means of employee participation.
More scientific rigor in intervention studies including obtaining accurate
baseline measures is needed. In particular, in order to be able to take
into consideration cause and effect, a random control design (RCT), is
recommended.
Whilst as far as possible observing principles for sound scientific
design, realism with respect to what is possible within the context of
organisations which are continuously experiencing change and where
participants are unlikely to remain as passive study objects.
2. Research method
The aims of this research were to:
devise and test the appropriateness as well as the effectiveness of a
risk assessment tool
develop, implement and evaluate three different bullying intervention
programmes
The study was carried out in five host organisations: These include:
1. Civil Service Department (London and North England)
2. NHS Mental Health Trust (South England)
3. Acute NHS Trust (North England)
4. Acute NHS Trust (South England)
5. Police Force (South England)
Although all five organisations were from the public sector, with the exception
of the two acute NHS trusts, they are all very different and therefore the
generalisability and transferability of our findings should not be affected.
In order to accomplish the aims of this research, a mixture of quantitative and
qualitative methods were applied. These are outlined below.
11
2.1 Focus groups
A focus group methodology was used at two different stages of the project.
The purpose for conducting focus groups during the initial stages of the
project within participating organisations included:
1. Contribution to the development of context-specific policies;
2. Contribution to the development of a risk-assessment tool; and
3. Ensuring that the intervention programmes were developed and carried
out with reference to the local context and to address issues identified
as important by participants.
Focus groups are seen as useful instruments due to their interactive qualities
and their ability to explore perceptions of individuals, and, in particular, the
subjective meaning individuals assign to events (Liefooghe & Olafsson, 1998).
Group interaction also has the potential to provide a range of views on the
issues under consideration (Kitzinger, 1994), a fact which is considered
invaluable when individual and context-specific meaning is being explored.
Bellenger et al. (1976) argue that when a number of people collectively
explore ideas by providing others with insight into their views and way of
thinking, the outcome of the interaction is more than the sum of its parts. The
‘snowball’ effect, in which a random comment may bring about a ‘chain-
reaction’ of responses, accounts for part of this synergetic effect. As
interaction was at the heart of the study, the primary role of the focus group
facilitator was to ensure that the discussions took their course. No attempt
was made to interfere in natural group processes.
In total, 55 focus groups were carried out between 01/10/2003 and 26/02/04
involving a total of 272 people. In most cases, approximately 7% of staff from
each unit were involved in this part of the study.
A second stage of focus groups were scheduled six months after the
interventions for the purpose of discussing changes in behaviour and in the
work environment following implementation of the interventions. Altogether
eight post-intervention focus groups were carried out in three organisations
2.2 Questionnaire survey
In order to evaluate the efficacy of our intervention programmes and the
validity of the risk-assessment tool, details of which are given in subsequent
sections, we developed a comprehensive questionnaire comprising of
questions related to experiences of bullying and negative behaviour, and
potential risk-factors of bullying. The questionnaire also contained measures
on mental health, sickness-absence, intention to quit, self-rated productivity,
job-satisfaction, individuals’ ‘psychological contract’ and a number of
demographic variables collected to make it possible to compare the
experience of different groups.
12
Some of the above sections comprised of existing validated instruments whilst
others such as the Bullying Risk Assessment Tool (BRAT) needed
development by the researchers (see section 4). The 22-item NAQ-R
(Einarsen & Hoel, 2001) was used to measure experiences of negative
behaviour and bullying. To assess the state of employee psychological
contracts we used Robinson’s (1996) 7-item measure of fulfilled employer
obligations and 2-item measure of met employment expectations. Finally,
mental health was measured using the 12-item General Health Questionnaire
(Goldberg & Williams, 1988).
As one of the main aims of this study was to devise and test the
appropriateness as well as the effectiveness of a risk assessment tool, we
specifically set out to develop a bespoke instrument for inclusion in our pre-
intervention questionnaire survey and took the opportunity of the post-
intervention survey to refine it further.
In order to identify any changes resulting from our interventions , the
questionnaire was distribute to the same pool of people twice - prior to the
delivery of the interventions and six months after the interventions.
2.3 Interventions
As indicated in the introduction, the intervention literature, not least with
respect to psychosocial hazards, advocates a context specific or local
approach responding to local needs and based on broad employee
involvement (e.g. Kompier et al., 1998; Giga et al., 2003). However, since our
main aim was to compare effectiveness of interventions across different
organisational settings, we concluded that it would not be feasible to compare
five completely different approaches from a scientific perspective. Such an
approach would also mitigate against generalisability as well as transferability
of findings to other contexts. However, although the interventions were based
on a common core approach, the local context was taken into consideration
and the programme was tailor-made to meet the need and reality of the
individual organisations by means of examples given and tasks set. On the
basis of a review of the literature and discussions with experts in the field we
identified the following three interventions: policy communication, stress
management training and negative behaviour awareness training. Each of
these interventions and their rational are outlined in section 5.
2.4 Research design
In order to avoid possible problems associated with research methodology,
such as selection bias and cross-contamination of data due to communication
between participants taking part in different interventions, the intervention
programme was implemented in similar or parallel departments or units by
means of a clustered randomised selection process. In each organisation five
groups were selected and assigned one of the following roles:
13
Figure 1: Research design
Group (1)
(Control grp)
Group (2) Group (3) Group (4) Group (5)
Policy
Communication
Policy
Communication
Policy
Communication
Policy
Communication
Stress
Management
Training
Negative
Behaviour
Awareness
Training
Stress
Management
Training AND
Negative
Behaviour
Awareness
Training
A sample of approximately 20-25 managers (or other employees) from each
group were invited to attend relevant training sessions. Group (1) served as a
control group and did not take part in any intervention, group (2) were invited
to a 30-minute policy communication session, group (3) were involved in a 3-
hour stress management training programme in addition to the policy
communication, group (4) were invited to a 3-hour negative behaviour training
session in addition to the policy communication session, and group (5) were
involved in a full day training session covering policy communication, stress
management and negative behaviour awareness. Although the researchers
ideally would have liked to increase the scope or length of training, initial
discussion with participating organisation suggested that any further time-
commitment was not feasible.
2.5 Local steering committees
From the start we encouraged the formation of local steering committees. The
overall aim of steering committees was to assist researchers in developing
and implementing the research programme locally. It was suggested that such
committees have a strong management presence, with representation of
occupational health practitioners and employees (unions/staff associations
where appropriate).
To ensure management commitment to the process we also required that
each participating organisation made a financial contribution to the study.
Furthermore, from the outset we emphasised that successful implementation
largely rested on continuous management support and warned against
potential impact of internal pressures, organisational change and influence of
competing initiatives.
In most organisations, a nominee from human resources served as the main
point of contact for the research team as well as a facilitator for the steering
group. Active support from senior management was encouraged from the very
early stages of the project.
14
3. Pre-intervention focus groups
The aims of the focus groups were to assist in the development of the risk
assessment tool as well as inform the development of specific interventions.
Apart from recording each focus group, participants were provided with a list
of negative behaviours taken from the revised Negative Acts Questionnaire
(NAQ-R)(Einarsen & Hoel, 2001), which is a validated instrument in its own
right and used extensively in bullying research. The one-page form also
requested simple demographic information and gave participants the
opportunity to list other negative behaviours specific to their organisation.
Information from both the interview transcriptions and feedback from
participants have been used in developing our risk assessment tool.
In total, 55 focus groups were scheduled between 01/10/2003 and 26/02/04
involving a total of 272 people. We piloted the exercise during the first two
sessions which involved 13 people from the Civil Service. The remaining
groups were scheduled as follows:
7 further groups within the Civil Service Department’s London sites
10 groups within the Acute NHS Trust (North)
9 groups within the Acute NHT Trust (South)
17 groups within the NHS Mental Health Trust
10 groups within the Police Force
A majority of the focus groups were carried out according to the protocol.
However, this process has not been without its problems, with some focus
groups cancelled due to organisations not being able to recruit sufficient
participants to make the groups viable. This was not least the case when
attempting to organise separate sessions for certain groups such as ethnic
minority staff. Whilst a majority of focus groups consisted of 3-6 people,
scheduling difficulties were also common, demanding maximum flexibility on
behalf of the facilitator. In one rather extreme example, due to
misunderstandings not less than 21 people attended the same session. In this
case the problem was resolved by turning what was meant to be a focus
group into a fact-finding group-meeting, obviously different from the nature of
the focus groups but still able to provide valuable input to the study. The
reasons for the low participation in some sessions differ from one organisation
to another. However, in general this was not due to a lack of interest on behalf
of prospective participants but instead difficulties in getting time off work from
what were considered already understaffed departments. It was particularly
difficult to engage employees who were based in units that were located some
distance apart. Although facilities have not been ideal for conducting focus
groups at these remote sites, we have in many cases insisted that some
sessions be held there in order not to exclude certain groups of employees.
The focus groups centred on two issues: assessing which negative behaviour
participant considered particularly difficult to handle or deal with and; what
factors, situations or antecedents contributed to negative behaviour and
bullying.
15
As suggested in the figure (2) below, results from focus groups highlighted
behaviour involving some degree of personal attack as most difficult to deal
with across organisational setting, with behaviours such as ‘being humiliated’
(item 2), ‘being ignored or excluded’ (item 6), ‘hints from others to quit your
job’ (item 10), and ‘personal allegations’ (item 17) scoring highly (3 refers to
very difficult to deal with and 0 refers to not at all difficult to deal with).
However, discrepancies exist both within and between organisations. For
example, certain negative behaviours such as ‘being ordered to work below
their level of competence’ (item 3) are seen as normal, unproblematic and for
some even welcomed at times, whereas as for others it is much more
problematical.
Figure 2: Difficult to deal with behaviours breakdown for each organisation
Focus group participants also identified behaviours that were not listed on the
NAQ(R) and which they found difficult to deal with. These were taken into
consideration in connection with development of the intervention programmes.
Other findings from the focus groups are highlighted in the next section.
4. Development of the Bullying Risk Assessment Tool (BRAT)
Aim: To devise and test the appropriateness as well as the effectiveness of a
risk assessment tool developed on the basis of the current study.
Risk assessment is a commonly used approach throughout industry to identify,
evaluate and control potential harmful effects of physical hazards. However, a
number of observers have advocated that a risk assessment strategy should
also be applied to psychosocial hazards such as occupational stress (e.g. Cox
et el., 2000) and bullying (Spurgeon, 2003). Central to such approaches is the
distinction between a hazard (something that has the potential to cause harm)
and risk (the probability that such a harm will occur). According to Spurgeon
risk-assessment should be part of a risk-management approach which would
incorporate the following elements or steps: 1) hazard identification, 2) risk
reduction and 3) control of risk. It is important to note that such an approach is
focused on reduction and control as opposed to elimination of risk because
many workplace hazards might not be altogether eradicated as they often
0.00
0.50
1.00
1.50
2.00
2.50
3.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Series1
Series2
Series3
Series4
Series5
16
represent part of normal activities and interactions in the workplace. In this
respect, the risk assessment process is focused on ‘estimating the probability
and the size of possible outcomes, and then evaluating the alternative
courses of action’ (Wharton, 1992 cited in Clarke & Cooper, 2004). As risk is
considered to be the product of the frequency and the consequences of
exposure to a hazard, it infers that hazards are likely to have severe
consequences.
Bullying can also be compared to what Turner (1978) refers to as ‘slow
accidents’, as the likelihood of bullying actually occurring may not always be
clear due to the uncertain effects of hazards (Clarke & Cooper, 2004). It is
also acknowledged that the intangible nature of psychosocial hazards, and
their subjective nature which is open to interpretation by individuals,
undermines the potential to measure or estimate the risk on a calibrated scale
as is the case to a far larger degree for physical hazards. Finally, as
highlighted in the stress literature, to be effective any risk assessment tool
needs to be context specific in the sense that it reflects the hazards within the
environment to which it is being applied (Cox et al, 2000; Giga, Cooper &
Faragher, 2003) and its success would, therefore, to a large extent be
contingent upon the degree of employee involvement and management
commitment to the process.
These factors need to be taken into consideration when devising a risk
assessment tool for it to be effectively be applied to workplace bullying.
Essentially such a tool would focus on the risk-factors for negative behaviour
associated with the experience of bullying. To identify what may be
considered general or global risk-factors a literature review was undertaken.
The review identified the following factors as commonly associated with
bullying:
• Excessive workload (Appelberg et al, 1961; Hoel & Cooper, 2000)
• Lack of control / lack of control of time (Zapf et al, 1996)
• Role conflict (Eianrsen et al, 1994; Vartia, 1996)
• Team-conflict (Hoel & Salin, 2003)
• Organisational politics (Kräkel, 1997; Salin, 2003)
• Organisational change (e.g. Neuman & Baron, 1998; Lewis & Sheehan,
2003)
• Perceived job insecurity/contingent workers (Knorz & Zapf, 1996;
Quinland, 1999)
• Organisational culture (Ashfort, 1994; Rayner et al., 2002)
However, the degree of empirical evidence in support of individual factors
varied, with for some factors contradicting emerging evidence. This
information was then related or merged with the result from the focus group
study outlined in the previous section. On this basis we developed a 26-item
risk-assessment tool which was then piloted as part of questionnaire
distributed to establish base-line measures across the participating
organisations (N=1033).
17
Although there was some evidence of the viability of our draft tool, the results
of the factors analysis of the pilot and a critical assessment of all items
making up the tool, suggested a need for further refinement. As a result a new
29-items instrument, the Bullying Risk Assessment Tool (BRAT) emerged. A
factor analysis (varimax rotation) yielded five factors. For a full overview of the
29 items instrument see Appendix B.
Extracts or quotations from the focus group discussions and examples of
items associated with each factors are provided below. (Please note that
some items making up the scales are worded positively and others negatively).
4.1 Factor structure
4.1.1 ‘Organisational fairness’
Sharing of work between colleagues was a common issue in a number of
organisations, particularly when in came to menial tasks:
“…My experience has been that you get the same people do them again
and again…they are the ones who do the extra bit and you’ll get other
people who will always find an excuse why they haven’t got the time to
do them.” NHS Ward Manager.
o I don’t know how to seek help within my organisation if I have a
problem with my line manager or supervisor
Grievance problems with managers were thought to be a particularly difficult
issue for some Civil Servants:
"Nobody knows where to go to if there’s something wrong. If you’ve
got a problem with your manager, nobody says “Right, you’ve got a
problem with your manager, you go to this person”. If you’ve got a
problem with your manager – tough! You know…there’s no guidelines,
there’s no checklist. If you say something like “Alright, I’ve got a
problem – I can’t relate to my manager”. Maybe I want to talk, your
manager’s a man or something – I want to talk to a female member of
staff. Where do I go?”
Other items:
o I feel my contribution to the organisation is recognised
o People in this organisation is not rewarded properly
4.1.2 ‘Team conflict’
o ‘Different professional groups don’t work well together within my
unit”
Conflict among team members was an issue raised in many groups.
Some participants emphasised that there was evidence of a lack of respect
between different professional groups. This was particularly apparent in the
18
NHS where nurses often experienced aggressive behaviour from doctors. The
quote below came from one nurse who was trying to make sense of a
Consultant’s behaviour by suggesting:
“Well, we get shouted at in theatre all the time by Consultants. And you
just realise that, it’s very impersonal. It’s nothing personal. It’s just that
they’re frustrated and it’s upsetting, but you don’t let it get in the way,
because the life of that person on the table, they are in charge of it, and
you just think “Right…he’s having a bad day. The life of the patient is
more important than your feelings”. We just get on with it.”
Other items:
o I don’t get on with some of my colleagues
o I find my colleagues to be cooperative
4.1.3 ‘Role conflict’
o I am not clear about how to carry out my job
Role conflict was another area frequently surfacing as suggested in the quote
below.
Other items:
o I face conflicting demands in my job
o My job description is clearly defined
4.1.4 ‘Workload’
o I feel that there isn’t enough hours in the day to complete my
work
Issues around workload were frequently raised by focus group participants:
A Clinical Manager exposed the constant pressure placed on them by the
demands of their job and in particular the struggle to schedule tasks within
a working day:
“I raised this issue with my line manager about that the amount and
expectation, that I will not be able to achieve the amount of work
that is passed down to my level, that I felt was unacceptable. I said
it’s impossible for me to do this, there were going to be things that
aren’t done and I have to prioritise what I can do and I can’t and the
answer he gave me back was that you’ll just have to do the best
you can. He said it’s what we all have to do, we have to do the best
we can and I said to him but I don’t think that’s good enough
because to me that is bullying, this is the Trust bullying me! I don’t
have 37 ½ hours written in my contract, it just says hours as
needed I forget the phrase but it says you know to do the job.”
19
Other items:
o Existing work pressure makes it difficult to take time off work
o Vacant positions are filled without any delay in my unit
4.1.5 ‘Leadership’
o I have confidence in my line-manager’s abilities
Leadership or lack of leadership was a frequently recurring theme:
“They don’t know how to react. All they know is that they come in, they
shut their door, they do their work, the jobs get allocated, the little mice
all run around and order everything and that’s it. But there’s no
interaction there. They don’t know how to interact with people because
nobody’s ever trained them to do that.” (Civil Service)
“I do think HR has become more removed from it, the emphasis is
moving much more, especially in this organisation which is relatively
new, moving towards line managers managing their staff and dealing
with situations but I’m not sure that a lot of people have got the
competencies and skills at the moment to do some of the things that
they’re asked to do or are asked to do and it’s all a little bit too much.
“ (Civil Service)
Other items:
o My line manager is sensitive to how I feel
o My line manager values constructive criticism
4.2 Factor reliability
Factor reliability was generally good (Chronbachs alpha of 0.57 – 0.81). The
lowest scale reliability emerged for the factor labelled ‘workload’. However, by
removing the item ‘my unit often makes use of temporary staff’, the alpha
increased to 0.62. It appears that this item might have a different meaning in
some of the organisations due to their particular practices regarding
temporary staff.
Table 1: BRAT factor reliability
Factor Scale reliability
Organisational fairness - OF 0.61
Team Conflict- TC 0.72
Role conflict – RC 0.65
Workload – WL 0.57
Leadership – LS 0.81
20
4.3 Validation of the BRAT
To validate the instruments a number of statistical tests were carried out. The
analysis revealed that the instrument can be considered valid in as much that
all factors, as well as the overall instrument, adequately emerged as
predictors of negative behaviour (measured by means of the revised Negative
Acts Questionnaire ((Einarsen & Hoel, 2001) and self-labelled bullying (in
response to a global definition of bullying) and with adverse consequences as
measured by means of mental health scores (General Health Questionnaire).
The NAQ-R is considered to measure two distinct factors, work-related
harassment and personal harassment (Einarsen & Hoel, 2001).
Table 2: Pearson correlations for BRAT scales
NAQ-R
Total
NAQ-R
WH
NAQ-R
PH
GHQ Bullying
BRAT – 29 item 0.581** 0.635** 0.450** 0.458** 0.419**
Org. fairness - OF 0.433** 0.463** 0.349** 0.315** 0.345**
Team Conflict- TC 0.467** 0.451** 0.386** 0.337** 0.416**
Role conflict – RC 0.406** 0.483** 0.278** 0.438** 0.203**
Workload – WL-6 0.296** 0.374** 0.183** 0.253** 0.168**
Workload – WL–5 0.304** 0.376** 0.194** 0.269** 0.166**
Leadership – LS 0.454** 0.478** 0.340** 0.315** 0.361**
** Correlation is significant at the 0.01 level
• All factors correlated significantly with total score on the NAQ-R, with
self labelled bullying and with total GHQ score
• Multiple regression indicated that all five risk factors are potentially
independent predictors of negative behaviour (R-sq=34.2%)
• Leadership followed by role conflict emerged as the single strongest
predictor of work-related harassment
• Leadership followed by team conflict emerged as the single strongest
predictor of personal harassment
• Multiple regression revealed that all factors predict work-related
harassment (R-sq=41.2%). Team conflict, leadership and
organisational fairness predict personal harassment (R-sq=22.3%)
• For self-labelled bullying – leadership, team-conflict, organisational
fairness and role conflict emerged as independent predictors (R-
sq=22.5%)
Results of the regression analysis are listed in Appendix C
It is worth noting that of those reporting themselves as having been bullied
within the last six months, 92% also reported having witnessed bullying talking
place within the same time period. This might have implications for application
of a risk-assessment tool as the far less sensitive issue of witnessing bullying
could to a large extent replace self-labelling whilst at the same time providing
similar information regarding predictors of bullying.
21
In order to assess the quality and potential effectiveness of the BRAT, the
scale’s test-retest reliability should be considered. However, one might
question whether it is feasible to successfully conclude on such an
instrument’s reliability.
Due to the sensitive nature of the issues under consideration, individuals are
likely to be influenced by daily events and there responses, therefore, could
be skewed. In other words, one might question to what extent risk-factors
could be expected to remain stable over time, especially where perceptions of
an individuals surroundings will at least to some extent fluctuate with task
requirement and ongoing dynamics of interpersonal interactions.
4.4 Application of the BRAT
It is important to consider applying the tool in its own right. In this respect it
was anticipated from the outset that the BRAT would be used in conjunction
with a more objective measure focussing on issues judged to be associated
with negative behaviour and bullying. The objectivity of this data refers to the
fact that it is taken from existing sources and is seen to be collated impartially
without any direct reference to bullying. For example, whether certain
changes have actually taken place or not, actual level of absenteeism
according to company records etc. For a further discussion of this, see section
7.
Although the risk-assessment tool can be used to identify individuals at risk of
bullying, the instrument is primarily aimed at establishing risk at a group-level.
In other words to what extent does some degree of consensus of risk exist at
a group level. This is important as one of the key outcomes of applying this
approach is to inform decision-making and highlight internal priorities. In this
respect organisational (or unit) action may be triggered by both overall scores,
relative scores on various factors or even on particular items. Returning to the
individual, where unit scores or risk is considered low whilst the scores of
particular individuals soar, one may conclude that there are other triggering or
contributing factors to bullying, possibly on a personal or one-to-one basis.
Such a conclusion would also require a different organisational response than
those cases where high risks are revealed at group level.
5. Identification and development of interventions
One of the aims of this research was: To develop and implement, according to
a particular research methodology, three different bullying intervention
programmes
This section provides details of the interventions and their rationale.
22
5.1 Policy communication
Whilst participating organisations are likely to have policies in place, the aims
of the policy communication intervention include raising awareness of the
organisation’s policy on bullying and the duty of organisational members in its
implementation. To ensure that the policies were in line with best practice in
the field, a set of guidelines were developed by the researchers on the basis
of close examination of a large number of such policies (see Appendix D).
Policies from each organisation were then carefully examined and compared
against it. Although differences were identified with respect to scope and
depth, all five organisations were seen to comply with our best practice
guidelines.
Rationale of intervention: It is envisaged that the individuals’ and others’
awareness of the presence of policies, the responsibility of managers with
regard to the implementation of policies and the potential consequences when
in breach of policy, would all impact on behaviour.
As the quote from a NHS Manager suggests the intention may not always be
to bully but the perception of the individual on the receiving end is important:
“It’s not about whether it happens or not sometimes, it’s about the
perception of the individual…when you listen to the staff, the victim,
and then they say this is how it made me feel, and you open your
policies and you think “gosh, yeah, it’s a perceptual thing!”
A lack of trust in grievance procedures was also highlighted on a number of
occasions:
“I just think, yeah, the organisation has a lot of responsibility and they
say a good job, they’ll have all the policies and they’ll say to you well
you can tick all these boxes and it is confidential but the absolute
reality is that it’s never confidential.”
The training includes:
• A statement of intent from senior managers highlighting the fact that
such behaviour will not be tolerated
• Outline of the managers / supervisors responsibility with regard to the
implementation of the policy and responsibility for challenging bullying
behaviour
• A definition of bullying and examples of bullying behaviour
• An overview of the complaints / grievance procedure and details of key
contact persons
5.2 Stress management training programme
This intervention follows the assumption that any negative and abusive
behaviour on the part of managers are frequently the result of work-overload
and the failure to deal with personal stress. The programme focuses on the
strain aspect of the stressor-strain relationship and the ability of individuals to
23
cope, and covers issues such as stress awareness and reduction. According
to some researchers, by becoming more aware of the relationship between
behaviour, personality, coping and stress outcomes, individuals may be better
able to cope with their stress. The aims of the stress management training
include raising awareness of stress and its impact on individuals and the
organisation, and developing manager / supervisor understanding of how to
manage their stress as well as the stress of people they are responsible for.
Rationale of intervention: by better controlling precursors of negative
behaviour resulting from stress, negative behaviour and bullying would be
reduced
Although bullying behaviour may not be accepted as part of an individuals’
character, a senior manager from the Police Force raises the possibility that
his behaviour could change adversely when under pressure:
“When I’m under pressure the aspiration is not to and I don’t know
whether I always control very well but I would think it would be a really
powerful person who under pressure kept so stable all the time. So
whether when I’m under pressure I change a little bit and maybe
become more prescriptive in what I want, which could be actually
construed as bullying, which I don’t think it probably would be because
I’m not of that ilk, but that is a possibility, but I try to stay calm but I
probably don’t always do it.”
The training includes:
• Defining stress
• Causes and consequences of stress
• Differing responses to stress
• How to identify stress in self and others
• Coping with stress (self) – problem focused coping, time management,
task prioritisation and exercise
• Managing stress in others
It is important to emphasise that the main aim of this intervention was to
assess whether stress management techniques have an effect on levels of
negative behaviour and presence of workplace bullying, rather than testing
the efficacy of stress management techniques in their own right.
5.3 Negative behaviour awareness training programme
Based on current knowledge and good practice guidelines in the area of
stress prevention and management which argue for researchers to target
interventions around organisational-specific issues, the purpose of this
intervention was to develop an evidence and needs-based programme to
tackle conflict and negative behaviour. Using feedback obtained from focus
groups and risk-assessment exercises, the aims of the negative behaviour
24
awareness training include raising awareness of negative behaviour and its
impact on individuals and the organisation, and developing a shared
understanding of what acceptable / unacceptable behaviour is within the
organisation.
Rationale of intervention: By raising awareness in a group context of negative
behaviour and bullying and providing participants with appropriate tools to
deal with difficult situations, bullying would be reduced.
Some extreme negative behaviours experienced and witnessed by a number
of people are carried out as a ‘right-of-passage’ or part of socialisation
processes and are seen as a bit of fun by perpetrators:
“Yes, well I had lots of practical jokes played on me in my earlier
years and some of them just weren’t funny, some of them were
indecent assaults by another name but they just thought it was quirt,
funny and all this sort of stuff…I had practical jokes played on me
which humiliated the life out of me and its that, it’s very personal, that’s
not a professional criticism, its about the person, I’m not just [NAME]
the Officer, it’s [NAME] the woman…
The training includes:
• Individual experiences of negative behaviour
• Definition of bullying and categories of bullying behaviour
• Evidence from previous research including effects on individual and
organisation
• Situations that cause bullying behaviour (organisation-specific
evidence from focus groups)
• Transactional analysis: how to develop skills for positive interaction
• A statement of intent from senior managers highlighting the fact that
such behaviour will not be tolerated
• Outline of the managers / supervisors responsibility to challenge
bullying behaviour
5.4 Implementation of interventions
The intervention programmes were all delivered by one person, who was an
independent and qualified trainer. A pilot of the interventions was carried out
in a non-participating NHS trust with the researchers present. On the basis of
our own observations and feedback from the participating managers (N=10)
each of the interventions were refined.
The interventions were evaluated by means of participant feedback obtained
at the end of the training session and six-months afterwards, comparison of
pre-intervention and post-intervention survey data and feedback from post-
intervention focus groups.
25
6. Questionnaire survey participants and sample background
6.1 Sample - study participants
We requested assistance from all five participating organisations in the
distribution of approximately 100 questionnaires to a random sample of
employees from each of the five intervention groups (approximately 500
questionnaires in total in each organisation) on two occasions – pre-
intervention for baseline measure purposes and then approximately 6 months
post-intervention for evaluation purposes. Questionnaires were sent out by
participating organisations accompanied by a cover letter from them outlining
the study, introducing the research team, and assuring confidentiality and
anonymity. A stamped addressed envelope was provided for questionnaires
to be returned directly to researchers at the University of Manchester.
In general the questionnaires were completed in an acceptable manner, with
little data missing for any variable. Of the 2505 questionnaires sent out during
the pre-intervention survey 1041 were returned to the researchers,
representing an overall response rate of 41.5 %. A total of 884 questionnaires
from the 2499 sent out during the post-intervention survey were returned to
the researchers, representing an overall response rate of 35.4%. Details of
both the pre-intervention and post-intervention questionnaire surveys are
presented in table (3) below.
Table 3: Survey sample
Organisation Total
sample
No. of
questionnaires
Returned
Response
rate (%)
T(0) 508 260 51.2 Civil Service Dept.
T(1) 499 249 50.0
T(0) 497 185 37.2 NHS Mental Health
Trust (NHS 1)
T(1) 500 162 32.4
T(0) 500 142 28.4 Acute NHS Trust –
North (NHS 2)
T(1) 500 160 32.0
T(0) 500 212 42.4 Acute NHS Trust –
South (NHS 3)
T(1) 500 139 27.8
T(0) 500 242 48.4 Police Force
T(1) 500 174 34.8
T(0) 2505 1041 41.6 Total Sample
T(1) 2499 884 35.4
26
In addition to distributing the questionnaire survey to employees,
organisations were asked to respond to the objective part of our risk
assessment tool by submitting information on each of the five groups
participating in this research.
Demographic details of the complete sample and their experiences of bullying
are presented below.
6.2 Sample background
The respondents were asked to ‘tick the box that best describes yourself or
your situation’.
6.2.1 Gender
Although there were substantially more female respondents, particularly within
NHS organisations, the proportion of female and male employees during both
surveys remained approximately the same. The gender breakdown within our
sample reflects the nature of organisations involved.
Table 4: Sample gender breakdown
Gender
(%)
Pre
Post
Int
Civil
Service
NHS 1 NHS 2 NHS 3 Police Total
Sample
T(0) 46.5 22.8 24.8 19.0 57.2 36.2 Male
T(1) 52.0 28.8 20.5 13.1 53.8 36.4
T(0) 53.5 77.2 75.2 81.0 42.8 63.8 Female
T(1) 48.0 71.2 79.5 86.9 46.2 63.6
Pre-intervention
Gender
Female
63.8%
Male
36.2%
Post-intervention
Gender
Female
63.6%
Male
36.4%
27
6.2.2 Age
The mean age of respondents was 43 years at both survey points. The
majority of participants in both the pre and post-intervention surveys came
from the 35-44 and 45-54 age groups on both occasions, although employees
from the Civil Service organisation were evidently younger. The sample did
not differ significantly between the two periods.
Table 5: Sample age breakdown
Age
(%)
Pre
Post
Int
Civil
Service
NHS 1 NHS 2 NHS 3 Police Total
Sample
T(0) 40.61 45.18 43.68 43.13 42.88 42.84 Mean
T(1) 41.07 46.45 43.02 44.69 43.17 43.39
T(0) 6.0 4.1 3.6 2.5 4.0 4.2 16-24
T(1) 6.9 2.0 4.1 3.1 2.4 4.0
T(0) 28.2 12.3 15.9 17.5 17.9 19.2 25-34
T(1) 27.9 9.3 14.3 15.3 18.2 18.1
T(0) 27.4 25.7 34.8 31.5 31.3 29.9 35-44
T(1) 26.2 26.5 36.7 24.4 30.9 28.8
T(0) 25.0 39.2 28.3 35.0 32.6 31.7 45-54
T(1) 22.7 43.7 29.3 41.2 33.9 32.9
T(0) 13.3 18.7 17.4 13.5 14.3 15.1 55-70
T(1) 16.3 18.5 15.6 16.0 14.5 16.2
0%
5%
10%
15%
20%
25%
30%
35%
Pre-intervention 4.2% 19.2% 29.9% 31.7% 15.1%
Post-intervention 4.0% 18.1% 28.8% 32.9% 16.2%
16-24 yrs 25-34 yrs 35-44 yrs 45-54 yrs 55-70 yrs
28
6.2.3 Ethnicity
Although percentages for specific ethnic categories were low at both measure
points, an above national average figure of more than one in ten respondents
were from a Black and Minority Ethnic (BME) background (assuming that the
‘other’ ethnicity group classify themselves as ‘non-white’).
Table 6: Sample ethnicity breakdown
Ethnicity Percent:
pre-intervention
Percent:
post-intervention
White 88.7 89.3
Non-white: 11.3 10.7
Mixed 1.0 0.7
Indian 1.5 1.4
Pakistani 0.4 0.2
Bangladeshi 0.3 0.2
Chinese 0.6 0.7
Other Asian 1.7 1.6
Caribbean 2.1 1.2
African 2.1 3.0
Other Black 0.4 0.2
Other 1.4 1.4
6.2.4 Contracted hours of work
The majority of respondents in both our surveys reported to working full-time.
Approximately one in five employees reported to working part-time hours
during the pre and post-intervention surveys.
79.7
81.8
20.3
18.2
0
10
20
30
40
50
60
70
80
90
Full-time % Part-time %
Pre-intervention -
Contracted hours
Post-intervention -
Contracted hours
29
6.2.5 Employment contract
As suggested in bar chart below, the vast majority of participants were
employed on permanent contracts. The numbers for the other categories were
relatively small.
95.3
93.6
2.5 2.8
1.2 1.5 1.1
2.1
0
10
20
30
40
50
60
70
80
90
100
Permanent % Fixed period or
task %
Seconded % Other %
Pre-intervention - Contracted hours
Post-intervention - Contracted hours
6.2.6 Job type
Almost half the sample came from ‘professional / managerial’ grades. The
sample did not differ significantly between the pre and post intervention
survey period.
Table 7: Sample job type breakdown
Job type Percent:
pre-intervention
Percent:
post-intervention
Clerical / admin 22.7 23.4
Professional / managerial 49.7 49.1
Specialist / technical 16.8 18.0
Other 10.9 9.4
6.2.7 Supervisory responsibilities
Almost half of the respondents suggested that their job did not involve any
managerial responsibilities whatsoever. Approximately the same number
reported having a supervisory or managerial role. In this respect the sample
remained relatively constant between the pre and post intervention survey
period.
30
Table 8: Sample supervisory responsibilities breakdown
Supervisory responsibilities Percent:
pre-intervention
Percent:
post-intervention
No supervisory responsibilities 46.7 45.9
Supervisor 22.0 21.1
Middle manager 18.6 19.0
Senior manager 6.1 6.8
Other 6.6 7.2
6.2.8 Length of time with organisation
A large number of respondents were employed within their current
organisations for more than ten years, reflecting a relatively stable
employment relationship. The figures remained almost constant over the
intervention period.
8.1
6.6
21.2 21.3
15.9
17.3
10.7
9.5
44.2
45.2
0
5
10
15
20
25
30
35
40
45
50
Under 1 year 1-3 yrs 4-6 yrs 7-9 yrs 10 or more yrs
Pre-intervention - Time with
organisation
Post-intervention - Time with
organisation
7. Objective data
In line with the intervention literature’s suggestions regarding methodological
progress, it was an expressed intention of the study protocol to develop a
procedure by which management could systematically measure certain
organisational indicators believed to be influenced by the presence of risk
factors of negative behaviour and bullying. Compared to self-reported data
obtained by means of questionnaires, such data can be considered more
objective in the sense that they are not open to cognitive processing to the
same degree but simply report on recorded organisational data.
With reference to bullying literature we developed a report form consisting of a
series of questions on measures such as absenteeism, turnover,
grievances/complaints and various forms of changes potentially taking place
31
within the organisation, e.g. restructuring, financial cutback and change of
managers. To provide an accurate account of these measures without any
potential interference from line-management, these forms were to be filled in
by the human resources department. An example of the form is provided in
Appendix E.
Unfortunately, despite informing the participating organisations about the
need to collect such data and the format by which this would be undertaken,
several organisations had great difficulties in obtaining the data in the form
required. In some cases data were only available at an organisational level
and not at the unit level needed to measure any potential impact of the
interventions. Thus, incomplete records on several measures particularly at
T1 (six months after the intervention) largely undermined the opportunity to
validate self-report measure. Similarly, incomplete records would also
undermine the opportunity to successfully apply these measurements to the
evaluation process. We report on the findings where we have complete data
from both the pre-intervention and post-intervention in section 10.
As we needed a uniform approach to the collection of ‘objective measures’ in
order to be able to compare data across organisations, we saw it as our
responsibility to identify the measures and develop the forms and procedures
needed to collate them in a comparable format. However, with the benefit of
hindsight it may be argued that management (and steering committees) in
participating organisations could have been more involved in the process, in
particular with the identification of valuable organisational measures as well
as reviewing internal data collection processes at an earlier stage to ensure
that they could comply with our requests. In other words, potential problems
could have been identified and rectified at an early stage by widening their
input to the process and by testing the procedure.
8. Experiences of bullying in the workplace: baseline measures
This section reports on the main baseline measures taken prior to
implementation of the intervention programme. For comparison purposes, we
also present findings from our previous BOHRF study (Hoel & Cooper, 2000).
The section reports only on the overall findings. For a further breakdown of
findings for different demographic groups, see Appendix F. This is particular
relevant since several measures applied in the current study were the same
as those used for the nationwide study reported in 2000.
8.1 Experience of bullying
Before answering questions associated with experience of bullying,
respondents were presented with the following statement/definition:
We define bullying as a situation where one or several individuals
persistently over a period of time perceive themselves to be on the
receiving end of negative actions from one or several persons, in a
situation where the target of bullying has difficulty in defending him or
32
herself against these actions. We will not refer to a one-off incident as
bullying.
‘Using the above definition, please state whether you have been bullied at
work over the last 6 months’.
A total of 13.6% of respondents report that they have been bullied within the
last six months. The persistency of their experience is illustrated by the table
below.
Table 9: Percentage of respondents reporting to being bullied
No Yes
(Total)
Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-
intervention
86.4 13.6 3.2 5.9 2.6 1.4 0.6
BOHRF 1 89.4 10.6 1.9 6.2 1.0 0.8 0.6
Pre-intervention
No
86.4%
Yes
13.6%
These figures suggest that a substantial proportion of employees in the
participating organisations considered themselves to being bullied, thus
entirely justifying the attention to the issue and current intervention study.
Moreover, compared to the nationwide study the overall exposure rate is
higher. Looking at exposure rates for individual organisations, it is interesting
to compare them with occupational/industry average figures in the BOHRF (I)
- study which were 9.9%, 10.6% and 12.1% for the Civil Service, the NHS and
the Police respectively. As we don’t know if our samples are representative of
their respective organisations or indeed of their occupation/industrial sector,
we cannot say that there has been a rise in exposure rates over the last five
year. However, what we can say is that the problem certainly appears to
affect a substantial part of the workforce and remains an important challenge
across organisational settings.
33
Table 10: Percentage of respondents from each organisation reporting to
being bullied - baseline
8.2 Witnesses of bullying
A total of 44% reported that they had observed or witnessed bullying taking
place within the last 6 months, indicating that nearly half of all respondents
would have either directly or indirectly experienced the problem.
Table 11: Percentage of respondents reporting to witnessing bullying in the
last 6 months
8.3 Past bullying
More than a quarter of the sample who said they weren’t being bullied at
present had been bullied in the last 5 years. This is mainly in line with findings
from the previous study.
Table 12: Percentage of respondents reporting to experiencing bullying in the
past 5 years
No Total
Yes
Yes,
very
rarely
Yes, now
and then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Civil Service 87.5 12.5 3.9 4.3 1.6 1.6 1.2
NHS 1 89.2 10.8 2.8 3.4 0.6 2.8 1.1
NHS 2 84.1 15.9 2.2 8.0 5.1 0.7 0
NHS 3 83.4 16.6 3.4 7.8 3.9 1.5 0
Police 86.9 13.1 3.0 6.8 2.5 0.4 0.4
No, Never
(%)
Total
Yes (%)
Yes, but
rarely (%)
Yes, now
and then %)
Yes, often
(%)
Pre-intervention 56.0 44.0 25.8 13.9 4.3
No (%) Yes (%)
Pre-intervention 72.3 27.7
BOHRF 1 75.3 24.7
34
8.4 Targets of bullying
8.4.1 Gender
A slightly greater proportion of women (13.9%) reported being bullied
compared to men (12.3%). However, this difference was not significant.
Table 13: Gender differences in exposure to bullying
Female (%) Male (%) Sig.
Pre-intervention 13.9 12.3 NS
BOHRF 1 11.4 9.9 NS
0%
2%
4%
6%
8%
10%
12%
14%
16%
Pre-intervention 13.9% 12.3%
BOHRF 1 11.4% 9.9%
Female Male
8.4.2 Age
With the exception of the youngest respondent group, only minor differences
emerged when comparing bullying for different age groups. However,
compared to the nationwide study the younger respondents reported less
bullying whilst those over fifty years of age appear to be more frequently
exposed.
Table 14: Age differences in exposure to bullying
16-24
(%)
25-34
(%)
35-44
(%)
45-54
(%)
55-70
(%)
Sig.
Pre-intervention 4.9 13.4 14.0 12.3 15.9 NS
BOHRF 1 14.0 10.5 13.0 12.0 8.8 NS
35
0
2
4
6
8
10
12
14
16
18
Pre-intervention 4.9 13.4 14 12.3 15.9
BOHRF 1 14 10.5 13 12 8.8
16-24 (%) 25-34 (%) 35-44 (%) 45-54 (%) 55-70 (%)
8.4.3 Full-time versus part-time
As in the nationwide study, respondents in full-time jobs were more likely to
be bullied than someone in a part-time job at a level bordering significance,
with 13.9% of full-timers as opposed to 8.9% of part-timers being bullied.
Table 15: Working hours differences in exposure to bullying
8.4.4 Ethnicity
Respondents of ethnic groups other than white were substantially more likely
to be bullied than white respondents, with 25.2% of non-white respondents as
opposed to 11.8% whites (p<0.001).
0%
5%
10%
15%
20%
25%
30%
Pre-intervention
BOHRF 1
Pre-intervention 11.8% 25.2%
BOHRF 1 10.5% 14.1%
White Non-white
Full-time (%) Part-time (%) Sig.
Pre-intervention 13.9 8.9 p=0.066
BOHRF 1 14.5 7.1 P<0.001
36
8.5 Organisational level of target
The current study confirms our previous findings (Hoel & Cooper, 2000) that
bullying is a problem across organisational hierarchies with just minor
differences emerging between the organisational levels or status groups.
Table 16: Organisational level of target
No
supervisory
responsibility
(%)
Spvsr.
(%)
Middle
Mgmnt.
(%)
Snr.
Mgmnt.
(%)
Other
(%)
Pre-intervention 13.5 13.0 12.6 11.9 18.0
BOHRF 1 9.6 9.1 10.6 8.5 11.0
8.6 Perpetrators of bullying
As evident from table 17 below, over 60% of bullying victims reported that
they were bullied by someone in a supervisory or managerial capacity. These
figures could suggest that there is a slight decrease in ‘vertical bullying’ since
our nationwide study. However, closer inspection of the nationwide study
showed that the NHS also had a lower percentage of vertical (top-down)
bullying (64.3%).
Table 17: Perpetrators of bullying
Table 18: Perpetrators of bullying: breakdown within organisations
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Pre-intervention 61.3 42.3 9.5 14.6
BOHRF 1 74.7 36.7 6.7 7.8
Spvsr /
manager(s)(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Civil Service 71.9 37.5 9.4 6.3
NHS 1 47.4 52.6 10.5 10.5
NHS 2 57.1 47.6 4.8 14.3
NHS 3 47.1 47.1 14.7 32.4
Police 77.4 32.3 6.5 6.5
37
8.7 How many were bullied?
Again the findings were very similar to the nationwide survey with just under a
third a third of respondents (29.9%) report to have been bullied on their own
or singled out for bullying, although substantial discrepancies exist between
the organisations. A total of 17.2% reported that everyone in their work group
were bullied, representing a minor increase compared to the BOHRF I study.
A breakdown of these findings for each organisation is provided in the table
19 below.
Table 19: How many were bullied
Table 20: How many were bullied: breakdown within organisations
8.8 Duration of the bullying experience
Approximately 60% of those who reported being bullied had been exposed to
bullying for more than a year, and approximately a third for more than 2 years.
Table 21: Duration of bullying experience
8.9 Experiences of negative behaviours
‘Someone withholding information which affects your performance’ followed
by ‘having your views and opinions ignored’ were the two most frequently
experienced negative behaviours. Amazingly, looking at the order by which
these behaviours were experienced it emerged that the ranking of the
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Pre-intervention 29.9 53.0 17.2
BOHRF 1 31.2 54.9 14.8
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Civil Service 31.3 46.9 21.9
NHS 1 38.9 55.6 5.6
NHS 2 15.0 55.0 30.0
NHS 3 32.4 58.8 8.8
Police 30.0 50.0 20.0
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Pre-intervention 20.0 21.5 25.9 32.6
BOHRF 1 16.8 16.4 27.5 39.3
38
behaviour was a near replica of the order emerging from the nationwide study
with only two behaviours changing place. This strongly suggests that although
organisational differences exist, both studies provide an accurate picture of
the nature of the behaviours experienced.
Table 22: Experiences of negative behaviour
*(R)egularl y or (O)ccasionally experiences
8.10 Outcomes of bullying
In line with previous research the correlations in table 23 below suggest that
bullying is relatively strongly correlated with mental health. Among other
outcomes often associated with bullying and with direct implications for the
organisation, the correlation with intension to quit is particularly strong. In both
cases the strength of correlations is substantially above those found in the
nationwide survey which were 0.267 and 0.212 respectively.
Table 23: Negative Acts: Breakdown within organisations
*
Pre-intervention (%) BOHRF 1 (%)
O 47.0 54.0 Someone withholding information which affects
your performance
R 10.0 13.3
O 40.4 49.3 Having your opinions and views ignored
R 8.2 7.8
O 33.8 39.3 Being exposed to an unmanageable workload
R 13.7 14.6
O 31.0 35.1 Being ordered to do work below your
competence
R 9.4 10.7
O 29.2 42.2 Being given tasks with unreasonable or
impossible targets or deadline
R 8.7 9.7
O 24.5 32.0 Having key areas of responsibility removed or
replaced with more trivial or unpleasant tasks
R 6.9 6.1
O 22.5 29.8 Spreading gossip
R 4.8 4.1
O 22.4 25.1 Being shouted at or being the target of
spontaneous anger R 4.7 4.7
O 22.3 27.8 Being humiliated or ridiculed in connection with
you work
R 4.6 3.6
Outcome measures
Have you been bullied at work
over the last 6 months?
p
Mental Health (GHQ 12) 0.392 <0.001
Intention to quit 0.328 <0.001
Job satisfaction -.178 <0.001
Psychological contract -.257 <0.001
39
9. Experiences of bullying in the workplace pre and post intervention
comparisons
By comparing baseline and post-intervention data, this section reports on the
main changes associated with bullying and negative behaviour which have
occurred within the duration of this study. For a breakdown of further pre- and
post-intervention, see Appendix G.
9.1 Experiences of bullying
During the baseline measure, a total of 13.6% of respondents reported to
being bullied in the previous six months. This measure increased slightly to
14.3% post-intervention (see table 24 below).
Table 24: Experiences of bullying – pre and post intervention
Pre-intervention
No
86.4%
Yes
13.6%
Post-intervention
No
85.7%
Yes
14.3%
As highlighted in table 25 below, experiences of bullying remained relatively
stable within all the organisations between the pre and post-intervention
period, except for a marked increase in experiences within NHS (3) from
16.6% to 23%, and a decrease in experiences in the Police Force from 13.1%
to 11.8%.
No Total
Yes
Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-intervention 86.4 13.6 3.2 5.9 2.6 1.4 0.6
Post-intervention 85.7 14.3 3.7 7.4 1.2 1.0 0.9
40
Table 25: Organisational breakdown of experiences of bullying – pre and post
intervention
9.2 Perpetrators
With a small reduction in numbers for supervisors / managers as perpetrators
of bullying, it is tempting to suggest that our interventions, which were
predominantly targeted at this group, could have had an effect on their
behaviour. However, the increase in numbers of colleagues who are bullying
(table 26 below) is more difficult to explain as is the small reduction in bullying
from clients / customers / patients.
Table 26: Perpetrators of bullying
As suggested in table 27 below, the reduction in bullying from managers /
supervisors was experienced in three of the five organisations involved, with a
marked decrease in NHS (2) (down to 45.8% from 57.1%) and the Police
Force (down to 65% from 77.4%). All organisations, except NHS (2) reported
a substantial reduction in experiences of bullying from clients.
No Total
Yes
Yes, very
rarely
Yes,
now and
then
Yes, several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Civil Service (T0) 87.5 12.5 3.9 4.3 1.6 1.6 1.2
Civil Service (T1) 87.4 12.6 3.6 4.5 1.2 1.2 2.0
NHS 1 (T0) 89.2 10.8 2.8 3.4 0.6 2.8 1.1
NHS 1 (T1) 89.2 10.8 4.5 3.2 1.9 0.6 0.6
NHS 2 (T0) 84.1 15.9 2.2 8.0 5.1 0.7 0
NHS 2 (T1) 84.3 15.7 2.0 11.8 0 2.0 0
NHS 3 (T0) 83.4 16.6 3.4 7.8 3.9 1.5 0
NHS 3 (T1) 77.0 23.0 7.4 13.3 0.7 0 1.5
Police (T0) 86.9 13.1 3.0 6.8 2.5 0.4 0.4
Police (T1) 88.2 11.8 1.8 7.1 1.8 1.2 0
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Pre-intervention 61.3 42.3 9.5 14.6
Post-intervention 57.4 47.5 9.8 9.0
41
Table 27 Organisational breakdown of perpetrators of bullying
9.3 How many were bullied?
As highlighted in table 28, the number of people who are singularly bullied has
increased from 29.9% pre-intervention to 36.1% post-intervention. At the
same time there is a substantial decrease in work group bullying from 17.2%
to 10.9%.
Table 28: How many were bullied
A breakdown of the data by organisations (table 29 below) also highlights
organisational variations in the number of people bullied, but with an overall
trend of a reduction in bullying of ‘everyone in the work group’.
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Civil Service (T0) 71.9 37.5 9.4 6.3
Civil Service (T1) 66.7 33.3 10 3.3
NHS 1 (T0) 47.4 52.6 10.5 10.5
NHS 1 (T1) 52.9 64.7 5.9 5.9
NHS 2 (T0) 57.1 47.6 4.8 14.3
NHS 2 (T1) 45.8 50.0 25.0 16.7
NHS 3 (T0) 47.1 47.1 14.7 32.4
NHS 3 (T1) 54.8 51.6 3.2 12.9
Police (T0) 77.4 32.3 6.5 6.5
Police (T1) 65.0 45.0 5.0 5.0
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Pre-intervention 29.9 53.0 17.2
Post-intervention 36.1 52.9 10.9
42
Table 29: Organisational breakdown of how many were bullied
9.4 Psychological contract
The term ‘psychological contract’ is being increasingly used in organisational
studies. It suggests that reciprocal exchanges enable individuals to control the
giving and taking that is involved in work settings without the development of
feelings of injustice. Employees observe their working relationships as a form
of social exchange by contributing effort and loyalty in return for discernible
benefits such as pay and recognition (Eisenberger, Huntingdon, Hutchinson &
Sowa, 1986). The ‘psychological contract’ is an implicit agreement between
the employer and employee that they will treat each other fairly. Although it is
not legally binding, both parties can aspire to maintain mutual trust and
potentially build a strong relationship. However, as it constitutes an emotional
bond, if it is broken it can lead to the feeling of betrayal and resentment
(Rousseau, 1995).
We used Sandra Robinson’s (1996) 7-item measure of psychological contract
fulfilment to evaluate the nature of employee psychological contracts. As
indicated in the table below (table 30) the figures suggest an overall negative
trend in psychological contract fulfilment in relation to experiences of bullying.
Table 30: The impact of experiences of bullying on victims’ psychological
contracts
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Civil Service (T0) 31.3 46.9 21.9
Civil Service (T1) 44.8 41.4 13.8
NHS 1 (T0) 38.9 55.6 5.6
NHS 1 (T1) 52.9 41.2 5.9
NHS 2 (T0) 15.0 55.0 30.0
NHS 2 (T1) 26.1 65.2 8.7
NHS 3 (T0) 32.4 58.8 8.8
NHS 3 (T1) 25.8 64.5 9.7
Police (T0) 30.0 50.0 20.0
Police (T1) 36.8 47.4 15.8
Mean
psychologic
al contract
score
Sample
mean
No Yes Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-
intervention
22.03 22.59 18.46 20.86 18.18 16.92 18.69 15.40
Post-
intervention
21.66 22.35 17.77 19.66 18.44 13.70 14.13 14.38
43
10. Efficacy of intervention: comparing baseline and post-intervention
measures for different interventions across organisations
In order to establish whether any changes have taken place in the period
between when base-line measures were obtained and six-months following
the interventions, a univariate analysis of variance was carried out on a
number of variables that could potentially be affected by the interventions.
These results are based upon an analysis of two sets of questionnaire
responses: Out of the 2,505 questionnaires distributed prior to the
interventions (baseline) a total of 1,041 were returned, representing a
response rate of 41.5%. 2,499 questionnaires were distributed for the post-
intervention (t1) measurement from which 884 were returned, representing a
reduced response rate of 35.4%. Overall these figures can be considered very
satisfactory for these types of studies.
The intervention groups referred to below are numbered as follows:
1 Control group
2 Policy communication (only)
3 Stress (and policy communication)
4 Negative behaviour awareness (and policy communication)
5 All (Stress, Negative behaviour awareness & policy communication)
10.1 Overall results
A series of analyses revealed that whilst a significant difference was found
between baseline scores and post-intervention scores for some variables
within all five groups, there was no significant difference for any of the key
variables between the experiment groups with respect to any change between
baseline and post-intervention scores. Overall scores for each intervention
group (scores across the five organisations) have been included in the tables
below and are introduced at the bottom of each table.
To examine the potential impact of different interventions in each of the
participating organisations, we carried out a univariate analysis of variance
(Test of between subjects effect). Overall no statistically different results
emerged for any of the key variables. Whilst in some cases there were
substantial differences between interventions, these differences were not
systematically upheld across all five organisations. It is, therefore, impossible
at this stage to conclude that any particular intervention or combination of
interventions is more effective than any other. However, in order to elucidate
trends in the data a closer examination of key variables was undertaken. The
data for each of these variables are presented and discussed below.
44
10.2 Bullying
Table 31: Bullied within the last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 baseline 1.21 1.27 1.20 1.41 1.42
T1 1.15 1.31 1.28 1.31 1.50
2 baseline 1.09 1.32 1.18 1.18 1.83
T1 1.24 1.17 1.38 1.06 1.23
3 baseline 1.50 1.17 1.25 1.24 1.65
T1 1.38 1.25 1.31 1.06 1.65
4 baseline 1.26 1.49 1.40 1.36 1.30
T1 1.35 1.31 1.29 1.69 1.45
5 baseline 1.29 1.36 1.15 1.34 1.22
T1 1.16 1.50 1.09 1.38 1.22
All orgs. Baseline 1.25 1.33 1.23 1.32 1.46
T1 1.23 1.31 1.27 1.31 1.43
Higher scores suggest increased levels of bullying
The above table indicates that very minor changes have taken place during
the six-month period between baseline and post-intervention measurements.
With more than 85% of respondents answering no (1) to this question, as well
as given the way the question is phrased which covers experience around the
time of the interventions this may be expected.
To be able to spot potential trends in the relative change between the base-
line and post-intervention (T1) measures, we simplified the above table by
means of showing increases in scores as +; decreases as -; and no changes
as 0. The use of plus (+) and minus (-) only refers to whether there is a
numerical increase (+) or decrease (-) in the scores for the instrument applied
and does not indicate whether the change is in the desired direction or not.
However, to avoid confusing the reader the desired direction is highlighted for
each table.
The number of pluses or minuses reflects the degree of change in a particular
direction, with each plus (+) or minus (-) representing a 5% change from the
baseline measure. In calculating the relative change between baseline and
post-intervention measures the minimum score on any scale is taken into
consideration. This re-scaling is done in order to avoid underestimating any
change.
However, due to the way bullying was measured with the large majority of
respondents answering ‘never’, the table is somewhat skewed. The
calculation of relative changes is, therefore, based on absolute values as
showed in the table. Each plus (+) and minus (-) represents an absolute
change of 0.1.
45
Table 32: Trends in experience of bullying
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 0 - 0
2
+ - ++ - - - - - - -
3
- 0 0 - 0
4
0 - - +++ +
5
- + 0 0 0
Higher scores (+) suggest an increase in levels of bullying
10.3 Witnessing bullying
It may be useful to look at equivalent scores for those who have witnessed
bullying as they would account for a much larger proportion of the
respondents than targets themselves and thus, possibly give a more accurate
picture of the situation.
Table 33: Witnessing bullying over the last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 baseline 1.49 1.67 1.48 1.51 1.47
T1 1.35 1.80 1.41 1.62 1.78
2 baseline 1.56 1.48 1.65 1.68 2.03
T1 1.55 1.47 1.51 1.50 1.42
3 baseline 2.00 1.66 1.63 1.80 1.90
T1 2.07 1.73 1.46 1.44 1.82
4 baseline 1.61 1.82 1.89 2.02 2.00
T1 1.92 1.69 1.81 1.78 1.92
5 baseline 1.67 1.60 1.54 1.43 1.43
T1 1.55 1.64 1.29 1.59 1.39
All orgs. Baseline 1.63 1.66 1.62 1.69 1.73
T1 1.63 1.69 1.49 1.59 1.67
Higher scores suggest an increase in witnessing bullying
To be able to picture any changes between base-line and post-intervention,
the same exercise as carried out for the previous variable also apply to the
variable ‘witnessing bullying’.
46
Table 34: Trends in witnessing bullying over the last six months
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
- + 0 + +++
2
0 0 - - - - - - - -
3
0 0 - - - - 0
4
+++ - 0 - - 0
5
- 0 - - + 0
Higher scores (+) suggest an increase in witnessing bullying
The above figures reinforce the view that no clear conclusion can be drawn
with regard to effectiveness of a particular intervention. However, with the
exception of organisation 1, there appears to be a slight downwards trend for
all groups which have received some training. Given that training and other
attention to workplace problems would tend to sensitise participants to the
problem, this overall slight reduction could signal a small reduction in levels of
bullying.
10.4 Negative behaviours
Table 35: Negative Acts Questionnaire (NAQ-R) - Minimum score on scale =
22
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 26.78 28.62 26.96 27.31 28.97
T1 26.82 27.83 28.55 27.10 29.94
2 Baseline 27.70 32.29 30.00 26.90 30.59
T1 27.07 30.69 33.28 28.47 28.08
3 Baseline 31.44 28.31 27.75 27.14 32.97
T1 31.88 28.81 28.30 26.79 28.29
4 Baseline 31.00 28.29 28.68 29.15 30.61
T1 31.22 28.48 27.38 29.36 32.10
5 Baseline 28.13 29.78 27.70 27.81 28.76
T1 28.17 30.03 27.19 27.81 29.41
All orgs. Baseline 28.57 29.29 28.13 27.76 30.25
T1 28.63 28.96 29.23 27.89 29.30
Higher scores suggest an increase in reported negative behaviour
Again, to be able to spot potential trends in relative change between baseline
and post-intervention (T1) measures we applied the same procedure as for
self-labelled bullying (see above).
47
Table 36: Trends in Negative Acts (NAQ-R) Minimum score on scale = 22
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 - - ++++++ 0 ++
2
- - - - - ++++++
+++
++++
++
- - - - - -
3
+ + ++ - - - - - - -
- -
4
0 0 - - - - 0 ++++
++
5
0 0 - 0 ++
Higher scores suggest an increase in reported negative behaviour
The above table indicates that in the case of negative behaviour also, no
systematic changes appear to have occurred between baseline measurement
and the post-intervention measurement. With only approximately 10-20% of
employees in the units trained (primarily supervisors / managers), it is of
course possible that the amount of training or number of people trained is
insufficient to affect behavioural patterns. Although the effect of training
generally is expected to be greatest straight after the training has taken place,
it is also of course possible that a six months delay between interventions and
measurements might have been too short for any changes to occur and be
noted. Alternatively, different interventions may in some cases have different
effects in different organisational settings, with other external or internal
circumstances playing a part. For example, organisation 1 experienced
unforeseen changes in the time between baseline and post-intervention
measures were obtained, with the possibility of redundancy affecting a
substantial number of employees taking part in the study.
10.5 The psychological contract
When we compare the psychological contract of individuals from the five
interventions groups, our data suggests that there is an overall reduction in
psychological contract fulfilment scores except for individuals who took part in
the full day training session (group 5). It was, therefore essential to check
whether these findings were generally upheld when looking at the impact of
individual interventions. The same procedure used for bullying and negative
behaviour was again applied.
48
21.69
21.96
22.02
23.12
21
20.88
21.19
21.96
22.9
21.19
20.5
21
21.5
22
22.5
23
23.5
Group 1 Group 2 Group 3 Group 4 Group 5
Pre-intervention Post-intervention
Table 37: Psychological contract – (Minimum score on scale = 7)
Intervention group mean scores
Organisation
Time of
assessment
1 2 3 4 5
1 Baseline 23.21 21.77 22.04 21.43 20.71
T1 22.57 21.38 21.34 22.27 19.64
2 Baseline 21.78 21.67 21.41 24.21 22.87
T1 20.79 21.78 20.62 24.25 22.17
3 Baseline 18.79 21.50 21.65 22.37 20.41
T1 17.95 20.13 21.14 23.89 22.44
4 Baseline 20.44 23.23 23.07 23.91 17.70
T1 20.05 21.35 24.21 21.94 20.50
5 Baseline 22.40 21.51 21.64 24.13 22.18
T1 21.48 21.23 22.00 22.52 21.74
All orgs. Baseline 21.70 21.96 22.02 23.15 20.86
T1 20.95 21.16 21.78 22.86 21.17
Higher scores suggest an improvement in the psychological contract
Table 38: Trends in Psychological contract (Minimum score on scale = 7)
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 0 + -
2
- 0 - - 0 0
3
- - 0 ++ +++
4
0 - - + - - +++++
5
- 0 0 - 0
Higher scores suggest an improvement in the psychological contract
49
Again it was impossible to trace any clear patterns in the data. Since
increased scores in this case represent an improvement of the psychological
contact, with a few exceptions improvement of the psychological contract was
associated with training in one form or another, with the most positive
development associated with the full day training (all training).
10.6 Mental health
Table 39: Mental health measure by GHQ-12 - Minimum score on scale = 12
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 24.15 25.07 24.29 24.29 24.26
T1 24.77 24.79 25.00 23.86 24.88
2 Baseline 24.37 26.42 25.69 23.87 25.72
T1 25.59 24.79 26.20 23.69 25.64
3 Baseline 26.23 23.64 26.83 24.93 27.69
T1 25.30 25.81 23.04 24.00 25.84
4 Baseline 26.00 24.40 25.49 24.31 25.68
T1 24.90 24.66 25.14 24.60 24.67
5 Baseline 25.37 24.87 23.74 22.68 24.46
T1 26.74 26.50 22.94 24.94 24.11
All orgs. Baseline 25.08 24.86 24.95 24.00 25.45
T1 25.52 25.26 24.62 24.21 25.06
Higher scores suggest worsening mental health
Table 40: Trends in the mental health measure by GHQ-12- Minimum score
on scale = 12.
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
+ 0 + 0 0
2
+ - - 0 0 0
3
- +++ - - - - - - - - -
4
- 0 0 0 -
5
++ ++ - ++++ 0
Higher scores suggest worsening mental health
Again the scores are difficult to interpret. Mental health scores could be the
product of a number of different factors. Experiencing negative behaviour and
bullying at work may only represent just one of a number of factors present
both within and outside the context of work.
50
10.7 Organisational outcomes, absenteeism and intention to leave
Absenteeism and turnover behaviour have been the primary focus of previous
research with regards to organisational outcomes associated with bullying.
Although bullying and negative behaviour have been found to be associated
with absenteeism, with the previous BOHRF study indicating a discrepancy in
terms of absenteeism of seven days annually between those who reported
being bullied and those neither bullied nor having witnessed bullying, research
has found a relatively weak link between these variables. Thus, a Norwegian
study reported that bullying only accounted for 1% of total absenteeism
(Einarsen & Raknes, 1992), a figure which increases to 2% in a Finnish study
of heath sector workers (Kivimaki, 2000), and 3% in a recent Swedish study
(Widmark et al., 2005). It therefore does not come as a surprise that it is
impossible to identify any particular trend in the current data.
Table 41: Self-reported absenteeism (time off work last six months)
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 3.87 5.06 5.15 5.84 3.77
T1 1.92 3.60 2.74 1.73 3.13
2 Baseline 2.18 5.21 3.58 2.71 2.74
T1 2.59 5.95 4.59 2.40 4.19
3 Baseline 1.01 3.83 4.42 3.09 3.83
T1 2.26 2.22 5.66 3.15 4.47
4 Baseline 4.55 2.85 2.03 4.37 3.86
T1 10.62 6.86 2.69 3.65 3.68
5 Baseline 5.01 3.25 5.38 5.95 3.96
T1 3.41 2.30 2.96 5.65 2.55
All orgs. Baseline 3.54 4.00 4.19 4.56 3.67
T1 3.72 4.01 3.61 3.25 3.56
A far stronger relationship has repeatedly been found between bullying and
intention to leave, or turnover behaviour (e.g. Unison, 1997; Rayner, 1999;
Hoel & Cooper, 2000).
As the scores in the above table are calculated as a sum of five separate
scores, it would not make sense to represent any changes as a percentage as
done previously in this section with other variables.
51
Table 42: Considered quitting last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 2.08 2.22 2.30 2.58 2.42
T1 2.19 2.34 2.26 2.52 2.81
2 Baseline 2.77 2.52 2.82 2.36 2.83
T1 2.90 2.76 2.55 2.13 2.32
3 Baseline 3.00 2.38 2.17 2.23 3.06
T1 2.96 2.81 2.63 2.34 2.53
4 Baseline 2.76 2.84 2.73 2.41 2.90
T1 2.60 2.55 2.68 2.55 2.91
5 Baseline 2.13 2.38 2.20 2.39 1.66
T1 2.31 2.53 2.35 2.42 1.82
All orgs. Baseline 2.46 2.46 2.45 2.42 2.49
T1 2.52 2.56 2.47 2.41 2.49
Minimum scale score = 1: Increased scores suggest increased intention to quit
Table 43: Trends in intervention and levels of turnover (data incomplete for
organisation 1 and 5).
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
Al
1
++ + 0 0 +++++
2
+ +++ - - - - - - - - - -
3
0 ++++ +++++
++
+ - - - - -
4
- - - - 0 - 0
5
+++ - - ++ 0 +
Increased scores suggest increased intention to quit
The above figures are difficult to interpret. On the one hand it appears to be
impossible to detect any particular pattern with regard to turnover if one
focuses on relative change. On the other hand, when looking at the post
intervention scores (T1), all units which have received training beyond policy
communication actually report levels of turnover below the organisational
average, which again could indicate a slight change in the right direction.
10.8 Satisfaction with the atmosphere at work
An improvement of the work atmosphere was considered a potential and
desired impact of our interventions. We therefore asked respondents to
consider the following question: “In comparison to six months ago, how
satisfied are you with the atmosphere at work”. Respondents were given three
answer alternatives:’ ‘more’, ‘less’ and ‘about the same’. As can be seen from
the table below, an overwhelming number of respondents considered the
52
work atmosphere to have got worse over the last six months. This might give
some support to the idea that taking part in training may sensitise and make
them more aware of workplace problems. Furthermore, training might raise
expectations which may lead to dissatisfaction when expected change is not
forthcoming.
Table 44: Satisfaction with atmosphere at work compared with six months ago
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline -0.02 -0.16 0.17 0.05 -0.25
T1 0.00 -0.14 -0.12 -0.14 -0.35
2 Baseline -0.05 -0.31 -0.03 0.10 -0.10
T1 -0.03 -0.24 -0.36 0.03 -0.04
3 Baseline -0.23 -0.03 -0.22 -0.17 -0.16
T1 -0.04 -0.33 -.046 -0.22 0.11
4 Baseline -0.30 -0.04 -0.55 -0.19 -0.39
T1 -0.24 -0.06 -0.52 -0.24 -0.64
5 Baseline -0.09 -0.15 -0.02 0.12 -0.04
T1 -0.38 -0.24 -.012 -0.27 -0.39
All orgs. Baseline -0.12 -0.13 -0.10 -0.01 -0.18
T1 -014 -0.20 -0.29 -0.17 -0.22
Increased scores suggest improved satisfaction with work atmosphere
As the scale used to measure atmosphere at work goes from –1 to +1 (with 0
representing no change), this can be considered to be a 2-point scale. Thus,
a change in score =0.2 represents a change of 10% (or 0.1 =5%).
Table 45: Trend in ‘satisfaction with atmosphere at work’
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 - - - - -
2
0 0 - - - 0 0
3
++ - - - - - 0 ++
4
0 0 0 0 - -
5
- - - 0 - - - - - - - -
Increased scores suggest improved satisfaction with work atmosphere
53
10.9 The effectiveness of intervention: assessing ‘objective’ measures
As an alternative measure of the potential efficacy of interventions, we asked
the organisations to provide some ‘objective data’. The response in returning
objective data was varied thus undermining the potential of an appropriate
comparison (see comments in section 7). However, the data was complete for
sickness absenteeism with the exception of one organisation, and for turnover
only two organisations provided complete records at both points of
measurement.
10.10 Intervention and levels of absenteeism
As absenteeism figures varied very substantially from organisation to
organisation and even between units within the same organisation, it makes
little sense to compare raw data or days of absenteeism directly. Instead we
will report on relative change within each intervention group. The three sets of
figures reported thus relate to a) absenteeism level in unit compared to
organisational average prior to intervention (baseline); b) absenteeism level in
unit compared to organisational average post intervention (T1); and c) the
relative change in absenteeism level for unit between baseline and T1 (before
and after intervention). NB: Absenteeism is not reported for organisation 1 as
data for two of the units were reported together and, thus render any
comparison between units impossible. Positive (+) and negative (-) scores
refer to unit levels scores above or below the mean score for the organisation
respectively.
Table 46: Absenteeism comparison in intervention groups
Absenteeism:
intervention group mean scores
Organisation Time of
assessment
1 2 3 4 5
2 Baseline +2.25 +4.52 +4.34 +1.94 +2.44
T1 - 1.70 +1.04 - 1.83 +0.70 - 2.17
Rel. change - 3.95 - 3.48 - 6.17 1.24 4.61
3 Baseline - 0.50 +0.90 +0.60 - 1.00 - 1.40
T1 +4.22 +9.12 +7.52 +7.22 +3.82
Rel. change + 4.72 + 8.22 + 6.92 + 8.22 +5.22
4 Baseline +4.67 +2.02 - 1.21 +0.60 +3.09
T1 +0.24 +1.45 - 0.12 +1.49 - 0.93
Rel. change - 4.43 - 0.57 + 1.09 + 0.89 - 4.02
5 Baseline +2.20 +3.20 - 0.50 +7.20 - 3.45
T1 - 1.50 - 1.15 +2.65 +7.10 - 1.30
Rel. change - 3.70 - 4.35 + 3.15 - 0.10 2.15
Closer scrutiny of the above figures reveals no clear patterns with regard to
change in absenteeism associated with any particular intervention.
As has been argued previously in this report, whilst the association between
bullying and negative behaviour, on the one side, and, absenteeism on the
other, is weak, a much stronger association is found for intention to leave.
54
Table 47: Intervention and levels of turnover (data incomplete for
organisation 1 and 5).
Turnover:
intervention group mean scores
Organisation Time of
assessment
1 2 3 4 5
2 Baseline
T1
Rel. change
3 Baseline - 2.80 - 0.30 - 1.30 +4.40 +0.10
T1 - 1.20 - 3.30 - 2.00 - 3.10 -2.30
Rel. change + 1.60 - 3.00 - 0.70 - 7.50 - 2.40
4 Baseline +0.61 +3.31 - 0.73 - 4.40 - 3.92
T1 - 0.76 +7.06 - 1.66 - 5.44 - 1.64
Rel. change - 1.37 + 3.75 - 0.93 - 1.04 + 2.28
Unfortunately, the data for turnover is difficult to interpret. Whilst most
experiments groups show a reduction in turnover, this trend is not consistent
across all experiment groups.
10.11 Other ‘objective’ measures
Only one organisation provided all requested data. However, in this case no
changes were recorded between baseline and post-intervention measurement
(T1) for any of the requested categories.
10.12 Total assessment of effectiveness of interventions
The previous sections have revealed that it is very difficult to trace particular
trends in the data when looking at different variables which could have been
influenced by the interventions. To enable us to consider all key variables
together with the aim of looking for potential trends across the data, we put all
the ‘relative representations’ by means of increases or decreases in variables
scores into one table (see table 48 below).
55
Table 48: Assessment of interventions (desired direction of change for each
variable is given in brackets)
O
R
G
IN
T
BU
(-)
WB
(-)
NAQ
(-)
GHQ
(-)
PC (+) ATM
(+)
TO
(-)
1 1 0 - 0 + 0 0 ++
2 1 + 0 - - + - 0 +
3 1 - 0 + - - + + 0
4 1 0 + + + 0 - 0 0 -
5 1 - - 0 + + - 0 + + +
1 2 0 + - - 0 0 0 +
2 2 - 0 - - - - - 0 0 + + +
3 2 0 0 + + + + - - - - +++ +
4 2 - - 0 0 - - 0 - - -
5 2 + 0 0 + + 0 0 - -
1 3 0 0 ++++++ + 0 - - - 0
2 3 + + - +++++++++ 0 - - - - - - -
3 3 0 - ++ - - - - - 0 - - +++ +
+ + +
4 3 - 0 - - - - 0 + 0 0
5 3 0 - - - - 0 - + +
1 4 - + 0 0 + - 0
2 4 - - ++++++ 0 0 0 - - -
3 4 - - - - - - ++ 0 +
4 4 +++ - - 0 0 - - 0 -
5 4 0 + 0 ++++ - - - - - 0
1 5 0 +++ ++ 0 - - +++ +
+
2 5 - - - - - - - - - - - - - - - - - - 0 0 0 - - - - -
3 5 0 0 - - - - - - - - - - - +++ + + - - - - -
4 5 + 0 ++++++ - +++++ - - 0
5 5 0 0 ++ + 0 - - - +
(BU=Bullying; WB=Witnessed bullying; NAQ=Negative behaviour;
GHQ=mental health; PC=Psychological contract; ATM=Atmosphere;
TO=turnover/decision to leave).
For nine (45%) of the 20 experiment groups (table 49), the changes are in the
desired direction for most variables, over and above those observed in the
control group. In all of these cases there were improvements with regard to
negative behaviour and bullying (which was the key objective of the
interventions) and an improvement (or at worst no change) with respect to the
56
other variables measured (i.e. all variables with the exception of atmosphere
and turnover, both of which were measured by one-item measures).
For three of the experiment groups (highlighted in darker grey) this trend was
upheld across all variables with very considerable improvement reported for a
number of them. It is important to notice that two of the groups represented
units that had received a full days training or taken part in all three workshops,
whilst the third group had received negative behaviour awareness training.
Although great care must be taken when interpreting these findings, they do
appear to indicate that the training might have had the intended effect, in at
least some cases.
11. Intervention feedback from the trainer and participants
11.1 The trainer’s perspective
As part of the evaluation process, the trainer was asked to report back on
each scheduled session on a specifically devised form. In addition to receiving
written feedback, the researchers conducted a two-hour interview with the
trainer to explore some of his perceptions around the impact of the
intervention programme in each of the five organisations. In total the trainer
carried out 40 distinct interventions.
The trainer reported varying degree of success. In particular, the training
groups appeared to vary with respect to group-dynamics (sometimes cross-
questioning rather than interacting) and their ability to engage with the training
content/message. In this respect the climate of the workshop would impact on
process, what may be discussed and the outcomes.
Looking at the three different workshops the trainer made the following
comments:
Policy communication:
Were only effective when a large number of relatively senior managers
attended the session.
Stress management:
The nature of the workshop meant that they often become too focused on
‘blaming the individual’. Furthermore, problems rather than solutions often
took centre stage when organisational issues were emphasised. The
workshops were often undermined by the fact that they focused on
organisational issues which participants felt they were unable to influence due
to lack of power/ seniority. However, there appears to be a consensus that a
lot of negative behaviour could be removed by reducing stress levels.
Negative behaviour awareness:
As most people do not admit to experiencing bullying behaviour, time is
needed for problem recognition to take place. The presence of macho
attitudes or fear of not being politically correct may undermine the ability to
learn. In general, the training was more effective in dealing with negative
57
behaviour than raising awareness of one’s own role in creating or contributing
to the problem. The transactional models introduced to deal with the problem
were considered effective only when participants were able to relate their own
experience directly to the model.
The following issues were highlighted by the trainer:
• The establishment of constructive dynamics and sufficient time is
needed for experiential learning to take place
• With regard to quantity and quality of training, a minimum of one day
appeared to be most effective. This is also evident from some of our
questionnaire findings, which suggests that individuals who attended
the full day session covering policy communication, stress
management and negative behaviour awareness training, gained the
most out of the interventions.
• One of the main objectives of this research was to implement bespoke
interventions that were developed with the benefit of local knowledge
and which provided examples that were relevant to the target audience.
Thus, for the learning to be effective the audience had to be able to
relate to the models presented.
• Groups need to be homogenous in that individuals attending sessions
should be from similar levels of the organisation to create a more
secure environment.
• Key people were not invited to training sessions or they did not attend
when invited: the training would be more effective with people from a
position of power or who were more able to influence their
circumstances.
• Several external factors may influence the effectiveness of the training
including substandard rooms, noisy environments and busy
organisations with frequent interruptions / individuals leaving
prematurely to attend to their work.
11.2 Feedback from participants
Participants were provided with a one-page evaluation form to complete at the
end of all stress management and negative behaviour awareness training
sessions. A total of 193 evaluation forms were collected from all five
organisations.
Three quarters of respondents responded positively in response to whether
the training challenged individuals to think in new ways.
58
Individuals also responded very positively to ‘How relevant was the session to
the programme aims’, ‘How interesting did you find the session’ and ‘How
would you rate the overall content of the session’ by rating each item as
‘good’ to ‘very good’ (1 =poor, 5 =very good).
In addition to this, participants were asked to write down the main thing they
had learnt from coming to the workshops. Their responses include:
• Self appraisal / reflection
• Awareness of different types of bullying behaviour
• To ensure that communication channels are kept open
• To control my stress levels better
• Prioritise
12. Post Intervention focus groups
Altogether eight post-intervention focus groups were carried out in three
organisations.
The aim of these focus groups was to qualitatively assess the effectiveness of
the interventions. For that reason the focus group discussions concentrated
on issues such as a) what are the key issues you remember from your
Di d the trai ni ng chal l enge you to
thi nk i n new and di fferent ways?
Yes:
75.10%
No:
24.90%
4.33
4.23
4.27
1
1.5
2
2.5
3
3.5
4
4.5
Relevance Interest Overall
rating
Mean participant
feedback scores
59
training and b) to what extent has the training changed your own or others’
thinking and behaviour with regards to bullying and negative behaviour.
The discussions indicate that specific training content is still remembered and
considered making sense six months after it was received. The following
issues were emphasised in the discussions:
“I think as managers we’re all aware of it, we’ve been on diversity
training, but that workshop was quite good in defining the different
aspects of bullying” (female manager, Civil Service).
“I think it reinforced for me how to look at different ways of managing
conflict in strained environments. For me it was a good tool to enhance
what I already learnt as a trainer…you’re just standing back and
allowing everybody to express themselves in whatever way is good for
them and then almost offering a mini-counselling service before you
get on with what you’re trying to achieve and that has certainly helped
me build a relationship with the people I work with” (female trainer,
Police Force).
In particular a number of participants highlighted the usefulness of the
transactional analysis approach which formed part of the Negative Awareness
workshops:
“The different models…child / adult relationships….for a piece of
information it was useful and it’s nice as an observing thing to say “oh
right, you’re behaving like a child and I’m in parent mode and we’re not
quite negotiating”…It allowed me to take one step back and reflect”
(male Officer, Police Force).
“It gave us an insight and awareness. You go back and see things in a
different light. I think it was the transactional analysis that you can talk
to someone as a parent and they can respond as a child and you need
to be on an adult to adult level all the time… I thought that was good”
(female administrator, NHS).
“I’ve used it (transactional analysis). I’m not saying how I’ve used it has
been successful in reflecting on my own performance. I’ve used it to try
and get a different outcome and sometimes it’s worked and sometimes
it hasn’t. You’ve got to be aware of the recipient of your message…just
may not want to know” (male Officer, Police Force).
Some people emphasised that the training had particularly helped them
improving their understanding of the problem:
Defining the problem
“I think as managers we’re all aware of it, we’ve been on diversity
training, but that workshop was quite good in defining the different
aspects of bullying” (female manager, Civil Service).
60
“I came away with was that it was to inform staff, colleagues etc of the
situation so that it was not considered as an overt activity, something
that was going on regularly…sort of making sure that those negative
traits and tendencies were managed better because some of them
were to improve productivity or certainly [make people] more motivated
with what they were doing. They needed to modify their approach so
that people weren’t put under undue pressure” (male, Civil Service).
Impact on behaviour
“I react differently to aggression now. Aggression is just self expression.
It may not be necessarily a personal attack on you but a lot of people
tend to deal with aggression with aggression. But I take myself out of
that and going in to that session reinforced what I need to and put into
place for my own protection before I actually react to a situation”
(female trainer, Police Force).
“It had made me aware that behaviours that I had seen could be
perceived as bullying, that a situation that I was managing someone I
needed to be mindful that my approach to a subordinate would need to
be necessarily open in ways which would project me coming into a
situation where someone can say “you’re bullying me”. No. Leaving the
door open at all points for someone to express, you know if they’ve got
reservations, and doing it in a manner if they did that they would feel
assured that we can find a way round that” (male, Civil Service).
“I think what we’re talking about is more broader than bullying and
harassment, and we’re talking about negative behaviours that we have
observed or experienced which of themselves may not be as painful as
individual instances of bullying but are equally relevant to the
department in terms of getting a happier more productive workforce.
And therefore to focus exclusively on bullying or harassment isn’t
terribly productive because I might sit there and think I wouldn’t bully
anybody or harass anybody and I have never had experience of those
things but I’ve had plenty experience of less painful negative
behaviours which I myself would like to learn from and would
encourage others to learn from. And I think this sort of drip feed will
then help them to reduce instances of bullying and harassment” (male,
Civil Service).
“It certainly is there in my mind because if the training that this is not
punitive action…I think that the one thing that the training will have
done is it will have usefully indicated to people that the perspective that
they will view their behaviour at with subordinates, with colleagues,
superiors etc, that’s something that needs to be analysed. It’s maybe
the hardest thing as individuals that we face is to examine ourselves as
best as we can from the outside…” (male, Civil Service).
61
Indirect suggestions that behaviour has changed due to interventions
(individuals more aware of their own behaviour)
“The Force is taking great steps to change things. The danger sometimes is
that it perhaps has gone too far…you’re very mindful that you’re not
bullying…There is a fine line between the definition of bullying and actually
having to tell someone that that needs to be done” (male Officer, Police
Force).
“we joke and muck around as friends would but at the same time we are
slightly uncomfortable with doing so because we feel quite heavy on our
shoulders, the responsibility to be diverse and not to intimidate anyone or
harass them or bully them and actually that can be negative because it
inhibits what in other circumstances would be normal interaction and so I find
myself saying “hold on, you shouldn’t be saying that sort of thing” (female,
Civil Service).
Uncertainty or suggestions that interventions may not have worked or
behaviour
“It’s a long haul process. I had a situation not so long ago where the
nurse actually turned around and said she didn’t want to take it further
because she felt that it wouldn’t get anywhere… if people feel like that
then it isn’t working very well” (female nurse manager, NHS)
“Well if people obviously have a problem then they will now know
where to go…the information is there if you need it, but I don’t think the
overall perception of the attitude of the management board has
changed…things have changed at the grass roots level” (female, Civil
Service)
“I think the Force has worked on it over the years. It certainly has not
eradicated it…I think the Force is putting forward a positive message
that bullying will not be accepted or tolerated. That has potentially had
the effect of making everybody know where we stand and perhaps
putting people off what may be perceived as bullying behaviour but on
the other hand it also has the effect of driving it underground and
making bullies more devious…I think people use less obvious ways of
wielding their power or influencing other people because they know
they face potential discipline if they get caught” (male Officer, Police
Force).
“Because I haven’t had any major issues within my team setting, we
haven’t had to test the processes so I don’t really know” (female nurse
manager, NHS).
“Impossible to say, I’ve changed jobs, different people.” (Male, Civil
Service).
62
What else can the organisation do?
Several participants emphasised the importance of ongoing initiatives and
training in order to bring about change.
“It was all relevant. Just having one session it’s difficult to enable it to
be a part of your life…It’s got to be on an on-going basis then it
strengthens you as a character to be able to deal with any conflict and
manage yourself better…It’s got to be on-going” (female trainer Police).
“…the training has had a vague but definitely instilling influence in me
and my approach” (male, Civil Service).
However, the answer is not necessarily associated with training per se but
with who receive the training. In other words, to be effective the training needs
to be targeted at those in need of training.
“We feel that we are…lower down the ladder when it comes to the
supervisory ladder, we think it should have gone up higher.” (female
administrator, Police Force).
“I suppose it comes down to how people are selected for training in the
first place” (female nurse, NHS).
“I think part of the problem is that you are preaching to the converted. I
don’t think we would be here if we were not aware of the bullying and
harassment as an issue anyway” (male, Civil Service).
“You’ve got a focus group of five people who’ve put themselves
forward. Probably people slightly on the goody side of normal, probably
the people that don’t need this {training]. The people who you do need
here are the people who didn’t put their hands up and aren’t here.”
(male, Civil Service).
From the focus groups discussions referred to above we can conclude that
although the training receives good marks and for some is considered to have
led to behavioural change, the picture is somewhat mixed.
63
13. Discussion
The present study confirms previous UK findings, suggesting that workplace
bullying is a serious workplace problem affecting a substantial number of the
workforce either directly or indirectly. The inclusion of some research
instruments or scales used in the nationwide study sponsored by BOHRF and
published in 2000 (Hoel & Cooper, 2000), has also made it possible to make
some comparisons with previous findings. Although an overall rate of bullying
of 13.6 and 14.3% at the two measurement times is somewhat higher than the
national average of 10.6% emerging from the study in 2000, most of the
current findings largely replicate those of the nationwide study and, thus give
strong support to the validity / authority of the larger scale study. Moreover,
with levels of bullying varying from 10.8 to 23.0% in the five organisations
taking part in the study, the identification and implementation of interventions
which could contribute to rectify the problem appears to be entirely justified as
well as timely.
As highlighted in the report, a risk-assessment instrument has successfully
been developed and tested. The validity of this instrument, the 29-item
Bullying Risk Assessment Tool (BRAT) was confirmed, with the BRAT (29
items) and each of its five factors, correlating highly with negative behaviour
and bullying as well as with mental health. Moreover, each of the five factors
independently predicted negative behaviour, whilst all factors with the
exception of ‘workload’ predicted self-labelled bullying. Although it is too early
to anticipate its full practical value, it is hoped that it will be tested out in other
studies in the near future. The BRAT represents the first attempt anywhere to
apply a risk-assessment approach to the issue of workplace bullying and as
such, must be considered a milestone. It might not yet have found its final
form but international interest in replicating or testing its qualities, particularly
from research teams in Belgium and Denmark, speaks for itself.
The study’s primary aim was to develop, implement and evaluate
interventions which alone or in combination could prove to be effective in
reducing negative behaviour at work and ultimately preventing workplace
bullying. To achieve this aim a highly complex and scientifically rigorous study
was designed and carried out. By involving five organisations from four very
different environments, with their own very specific structures, cultures and
routines, the researchers have had to juggle a multitude of issues in order to
carry out the study according to the original protocol and to ensure that
sufficient information was collated at various points. With this in mind we can
conclude that we have reached our overall aims and objectives.
The interventions applied in the study were all based on sound theory and
current understanding of bullying and its antecedents, and we believe that
they have adequately addressed the problem in order to bring about the
envisaged behavioural change. Similarly, the professionalism of the trainer
and the effectiveness with which the workshops were delivered were never in
question, with feedback from participant also appearing to give strong support
to this. However, with reference to the stress literature, locally developed and,
64
thus, fully contextualised training programmes developed in response to
particular local needs could be anticipated to have an even greater impact.
But the scientific design which aimed to compare interventions across
different settings, and thus allow for conclusions with regards to
generalisability and transferability of findings to be made militated against
such an approach.
Unfortunately, we have only to a limited extent been able to utilise the
objective measure which were obtained during the study. A number of factors
accounts for this shortcoming: Firstly; some organisations were unable to
obtain the data as requested (and agreed) due to difficulties with external data
processing mechanisms and incompatible formats where intervention units
were different from internal units of measurement. Secondly; organisations
were unable to obtain information about particular issues, e.g. grievance and
complaints records. With hindsight, such shortcomings could possibly have
been resolved by involving the organisations in developing the actual
measures as well as testing the system prior to the study.
Despite strong indications from participants that the training was relevant and
at least to some extent has had the intended behavioural impact as witnessed
by focus group accounts, there is insufficient evidence in the data to make
any conclusions with regard to the efficacy of particular interventions. This is
unfortunate, albeit not surprising, and seems to be in line with most
organisational intervention research (Murphy & Sauter, 2003). However, when
comparing the data across a number of variables, it is apparent that some
improvement in the desired direction for negative behaviour and bullying has
taken place in approximately a third of experiment groups, but with different
interventions associated with positive outcomes in different organisations. If
we consider these results with the feedback from participants on the day of
training and six months following the training, the trainer’s reports on delivery
and anticipated effectiveness, as well as focus group participant observations
and responses post intervention, we can tentatively conclude that the training
does seem to have at least some effect, although different interventions may
work better within different contexts. Moreover, for two of the experiment
groups all scores on the eight key variables were in the desired direction, with
a very substantial improvement noted for some of the variables. In both cases
these results have emerged from units where the experiment group have
received a full day’s training. The fact that this finding was not upheld across
all five organisations can be due to a variety of reasons, with the trainer’s
report in one particular case suggesting considerable unwillingness/resistance
on behalf of participants to engage with the training, and another,
experiencing threat of redundancies at the time of post-intervention
measurement.
65
To account for the overall results a number of factors need to be considered:
Intervention may increase awareness and expectations, and in
turn lead to dissatisfaction, affecting post-intervention
measurements.
It should be emphasised that raised levels of problem awareness
during the initial stages in a change process should not be considered
a problem per se and may even represent an asset. However, it is well
known from other studies (e.g. stress literature) that interventions might
contribute to sensitisation of a problem with higher scores on particular
variables as a result. In a similar way, interventions may also raise
expectations which subsequently could manifest itself in dissatisfaction
when change does not occur or when the pace of change is considered
to be too slow. For example, it is possible that by identifying and raising
awareness about stressors present in the work environment but at the
same time failing to, or being unable to provide participants with
necessary powers to address these stressors, further stress and
frustration may emanate. Following this line of argument it might,
therefore, be suggested that where no change has occurred between
the two measurement points, an actual improvement could have taken
place. The eight percentage point drop in response rate from
measurement one to measurement two might also has impacted on
results in a similar way, reinforcing the above argument. Thus, it is well
known from previous studies (e.g. Einarsen et al., 1994) that targets of
bullying are somewhat more likely (Einarsen and colleagues found a
difference of 10%) to respond that those who did not consider
themselves to be bullied.
The right people may not have been trained
As previously outlined, in order to have the maximum effect, our
intention was to aim the training at people in line management
positions. Although there is no problem in including employees without
supervisory responsibility in the training, where this has lead to line
managers being replaced by rank and file employees, it might have
reduced the maximum potential impact of the training.
There is some evidence to suggest that individuals who were
considered (by their managers or their colleagues) as potential
beneficiaries of the training, were actually not selected for training.
Moreover, there is also anecdotal evidence that some actually
deliberately avoided taking part in the training despite being selected,
by not turning up at the session or by nominating colleagues in their
place. Insufficient attention to planning and scheduling on behalf of the
participating organisations has in some instances resulted in low
participation rates and in other cases, people have left in the middle of
a focus group or a training session because of other work
66
commitments. Even with forewarning, participating individuals &
organisations have had to come to a decision as to whether to proceed
with crucial day-to-day work tasks, such as operating on patients, or to
attend focus groups / training sessions. It should be noted that
resource limitations on behalf of the researchers have meant that the
interventions have had to been planned on two consecutive days within
each organisation. Although this has worked well for research
purposes, in has caused organisational problems and possibly
contributed to lower participation rates. In principle the training could
have been split into different modules and could be delivered more
flexibly over a longer period of time.
By contrast, there is evidence to suggest that management have
consciously or subconsciously made a decision to apply training to
particular ‘hot spots’ or problem areas. Whilst this might contribute to
inflating bullying figures and other measurements, suggesting that the
data might not be representative of the entire organisation, such
decisions make complete sense and should in principle not affect the
analysis where the emphasis has been on relative change between
post-intervention (T1) and baseline measures.
Things are going on within the organisations impacting on
processes and individual outcomes
The post-intervention focus groups bear witness of the presence of
other factors potentially influencing findings. For example, in some
cases other initiatives aimed at organisational change, of which some
were also associated with training of participants, were going on at the
same time or in parallel with our project. Moreover, other organisational
processes may interfere with or even counteract our effort to bring
about change, e.g. restructuring or organisational change. For example,
in one of our organisations a reorganisation unknown to us at the offset
of the project, not only led to internal turmoil but also the potential of a
large number of redundancies.
The units of comparison (referred to as intervention group 1-5)
may not have remained the same during the entire course of the
study.
There is evidence that some individuals have been transferred to other
jobs or other units or may have left the organisation altogether. Others
may have been recruited during the period following the intervention.
However, this is to be expected and is likely to have only marginally
affected the results. Furthermore, in at least one of the organisations a
large number of employees from one of the experimental units were
moved to another unit in the middle of the study. Taking turnover rates
into account (approximately 10 percent), such factors would undermine
the value of straightforward comparison of pre and post intervention
measures.
67
Overall ‘critical-mass’ may not have been achieved.
Although we were up against the reality of limited resources and
commitment of the participating organisations, it may be argued that
too few people within each unit were trained in order to have any
measurable effect in the desired direction. Similarly, the amount of
training each individual received may have been insufficient for any
behavioural change to occur. In this case both factors are likely to have
influenced outcomes. The fact that participants selected for training
frequently did not turn up or, in some cases had to leave in the middle
of the training further highlight this problem. However, as emphasised
in the previous point, ‘critical mass’ may be as much related to the
total number of people trained as well as the amount of training
participants received.
The time between interventions and post-intervention
measurement (T1) is too short for any effects to have occurred.
For most intervention studies effects appear to be the greatest
immediately after the training where the content is relatively fresh in
the minds of participants, and then gradually diminishes over time,
especially where the training is not repeated (Giga, Cooper, Faragher,
2003). However, the nature of the problem of destructive behaviour
and bullying and its association with the culture of the unit and the
wider organisation could suggest that change would be slow and
possibly a delayed process. However, the relatively short period
between measurements could not to be avoided due to the total
duration of the study and the expressed wishes of the participating
organisations. To test this assumption a second measurement (T2), i.e.
12 months after the intervention is currently being discussed with one
of the organisations. It is questionable whether a T2 measurement
would necessarily bring about greater clarity bearing in mind some of
the problems discussed above, which may to some extent have had an
affect on this organisation.
The above discussion reviews a number of factors which alone or in
combination may help to explain why it was not possible to trace any clear
pattern in the quantitative data and, thus, prevent us from drawing any firm
conclusions with regard to the effectiveness of any particular intervention or
combination of interventions. However, looking at the complete evidence
presented, the combination of quantitative and qualitative evidence appears to
provide some evidence that the intervention may have had a small positive
effect, particularly for those attending the full day training.
Whilst it might be somewhat disappointing not to be able to draw any clear
conclusions with regard to the efficacy of applied interventions, this study, the
first of its kind with respect to bullying and destructive behaviour, has provided
us with considerable insight into organisational intervention processes, and
which therefore is likely to greatly benefit future studies. It is to such learning
points we now turn our attention.
68
14. Conclusions
As far as we are aware, this report completes the very first scientific study into
evaluating the effectiveness of management interventions for tackling
interpersonal conflict in the workplace. Despite many obstacles and
challenges on the way, we are confident that we have met our original aims
and objectives in connection with this ambitious study. A risk assessment tool
(BRAT) to assess the risk of negative behaviour and bullying has been
successfully developed and validated. In our opinion, the BRAT has
considerable scope for practical application.
By means of a complex study design, and by applying rigorous scientific
standards throughout, we developed, implemented and evaluated three
different interventions, all theoretically sound and based on a review of the
literature and local contextual information. In order to assess the effectiveness
of the interventions or combinations of interventions, we applied a randomised
controlled trial (RCT) design and used a wide range of data of a quantitative
as well as quantitative nature in the validation/evaluation process. Five large
organisations, within the public sector took part in the study and the same
design was adopted for all organisations to allow for evaluation purposes and
generalisation of findings.
With regard to the survey data obtained prior to, and six months after the
interventions, no statistical differences emerged for any key variable between
intervention groups across the five organisations. Thus, it was impossible to
establish the efficacy of particular interventions or, combinations of
interventions. However, when all outcome variables were considered together,
an improvement or change in the desired direction appears to have taken
place in some of the intervention groups, particularly where participants had
taken part in a full day’s training and, thus, received the complete training
programme. These findings were also supported to some degree by feedback
of a qualitative and quantitative nature from the trainer, trainees and
participants of post-intervention focus groups, whether they had previously
had taken part in the training or not. When the same outcomes were not
reported across the five organisations, various factors, some of which were of
an organisational nature and could be accounted for, might have contributed
to this result. Firstly, although it was our intension to primarily focus the
intervention on managers or people with supervisory responsibility in order to
achieve the maximum impact with limited resources, there is clear evidence
that in many cases this was not the case, with employees without supervisory
responsibility, replacing members of the key target groups or, indeed,
individual managers or supervisors identified as in need of training. Secondly,
and related to the previous point, with only 5-20% of staff within a
measurement unit being trained, insufficient numbers or a ‘critical mass’ was
not achieved to bring about a measurable effect. Thirdly, other factors of an
organisational nature or the amount of training provided may have been
insufficient for experiential learning to take place. Fourthly, it is well known
from other intervention studies that training or other interventions might
sensitise participants and the environment within which the interventions are
69
taking place to the problems, with greater awareness manifesting itself in the
shorter term in dissatisfaction and inflated scores on certain variables.
With regard to the apparent success of the full day’s training combining all
three interventions, this could have as much to do with the amount of training
as the particular nature of the training. Thus, we will second the view of the
trainer that in order to achieve the desired change in behaviour, experiential
learning must take place. For a sensitive issue such as the one in question,
this would require positive group dynamics and the development of a safe
learning environment, both of which in most cases take time to establish.
The present study has provided us with many important learning points,
particularly with regard to the intervention process which we believe will
greatly benefit future intervention studies in the area of workplace bullying,
destructive behaviour and beyond. In this respect we would emphasise the
importance of continuous management commitment to the process, based on
shared in-depth understanding of the aims of the interventions, responsibility
for resource provision, resource allocation and, implementation of the entire
research project. However, to ensure a successful outcome wide participation
and involvement must be established from onset, allowing for interventions to
be fully contextualised or anchored in the local context. In this respect it is
tempting to suggest that the scientific rigour and RCT design applied to the
present study might in itself has militated against a better or more clear-cut
result. Thus, we have previously been warned that applying rigorous scientific
design when undertaking research in busy, rapidly changing environments is
fraught with difficulties (Kompier, 1999). Therefore, whilst a case-study design
where the interventions are designed to address the needs of the organisation
and based on wide employee involvement would make it more difficult to
make predictions with regard to generalisability and, indeed, transferability of
findings or tested approach, a case study approach might have been more
successful in establishing the efficacy of interventions.
Finally, our results suggest that there appears to be no ‘quick fix’ solution to
alleviate an organisation of the problems of destructive behaviour and bullying
or to ensure that bullying does not manifest in the future. In our opinion, the
only viable option open to the organisation is to embark upon cultural change,
a process that can be long and difficult and which would require continuous
management commitment. But as many organisations come up against
restricted resources despite their best intentions, it is important not to judge all
interventions within a limited scope or timeframe as a dissipation of resources,
or worse, as counterproductive. Thus, carefully planned and targeted
interventions based on sound theory could bring about change and may
represent a first important step on the ladder of cultural change, even though
definite evidence of any desired impact might be difficult to measure,
particularly in the short-term.
70
15. References
Appelberg, K., Romanov, K., Honlasalo, M., & Koskenvuo, M. (1991)
Interpersonal conflicts at work and psychosocial characteristics of employees.
Social Science Medicine, 32, 1051-1056.
Ashforth, B. (1984) Petty tyranny in organizations. Human Relations, 47 (7),
755-778.
Bellenger, D.N., Bernhardt, K.L. & Goldstucker, J .L. (1976) Qualitative
Research Technique: Focus group interviews. Qualitative Research in
Marketing. Chicago: American Marketing Association.
Clarke, S. & Cooper, C.L. (2004) Managing the Risk of Workplace Stress.
London: Routledge
Cooper, C.L. & Cartwright, S. (1997) An intervention strategy for workplace
stress. Journal of Psychosomatic Research, 43, 7-16.
Cox, T., Griffith, A. & Rial-Gonzalez, E. (2000) Research on Work-related
stress. Luxembourg: European Agency for Safety and Health at Work.
Einarsen, S. and Hoel, H. (2001) The validity and development of the Revised
Negative Acts Questionnaire. Paper presented to the 10
th
European Congress
of Work and Organizational Psychology, Prague, 16-19 May.
Einarsen, S., & Raknes, B. I. (1991). Mobbing i arbeidslivet. En undersokelse
av forekomst og helsemessige konseqvenser av mobbing pa norske
arbeidsplasser. (Bullying at work: a survey of frequencies and health
consequences of bullying at Norwegian workplaces). Forskningssenter for
Arbeidsmilijø: Universitetet Bergen, Norway
Einarsen, S., Hoel, H., Zapf, D., & Cooper, C. L. (Eds.) (2003). Bullying and
Emotional Abuse in the Workplace. International perspectives in research and
practice. London/New York: Taylor and Francis.
Eisenberger, R., Huntingdon, R., Hutchinson, S. & Sowa, D. (1986) Perceived
Organizational Support. Journal of Applied Psychology, 71(3): 500-507.
Giga, S. I., Cooper, C. L. & Faragher, B. (2003) The Development of a
Framework for a Comprehensive Approach to Stress Management
Interventions at Work, International Journal of Stress Management. Special
Issue: Stress and Its Management in Occupational Settings, 10(4): 280-296.
Goldberg & Williams (1988) A User’s Guide to the General Health
Questionnaire. Windsor: NFER-NELSON
71
Hoel, H. & Cooper, C.L. (2000) Destructive Conflict and Bullying at Work,
Manchester School of Management, University of Manchester Institute of
Science and Technology
Hoel, H. & Salin, D. (2003) Organisational antecedents of workplace bullying.
In S.Einarsen, H.Hoel, D.Zapf & C.L.Cooper (Eds.) Bullying and Emotional
Abuse in the Workplace. International perspectives in research and practice
(203-218). London/New York: Taylor and Francis.
Hoel, H., Cooper, C. L., & Faragher, B. (2001). The experience of bullying in
Great Britain: The impact of organisational status. European Journal of Work
and Organizational Psychology, 10, 443-465.
Hoel, H., Faragher, B. and Cooper, C.L. (2004) Bullying is detrimental to
health, but all bullying behaviours are not equally damaging. British Journal of
Guidance and Counselling, 32, 3, 367-387.
Hubert, A.B. (2003) To prevent and overcome undesirable interaction: A
systematic approach model. In S. Einarsen, H. Hoel, D. Zapf & C.L.Cooper
(Eds.) Bullying and Emotional Abuse in the Workplace. International
perspectives in research and practice (299-311). London/New York: Taylor
and Francis.
Kitzinger, J . (1994) The methodology of focus groups: The importance of
interaction between research participants. Sociology of Health and Illness, 16,
(1), 103-121.
Kivimäki, M., Elovainio, M., & Vahtera., J . (2000) Workplace bullying and
sickness absence in hospital staff. Occupational and Environmental Medicine,
57, 656-660.
Knorz, C. & Zapf, D. (1996) Mobbing – eine extreme Form sozialer Stressoren
am Arbeitsplatz. Zeitschrift für Arbeits- & Organisationspsychologie, 40, 12-21
Kompier, M., Geurts, S., Grundemann, R., Vink, P., & Mulders, P. (1998)
Cases in stress prevention: The success of a participative and stepwise
approach. Stress Medicine, 14, 155-168.
Kräkel, M. (1997) Rent-seeking in Organisationen - eine ökonomische
Analyse sozial schädlichen Verhaltens. Schmalenbachs Zeitschrift für
Betriebswirtschaftliche Forschung, 49 (6), 535-555.
Liefooghe, A.P.D., & Mackenzie Davey, K., 2001, Accounts of workplace
bullying: the role of the organization. European Journal of Work and
Orgniazational Psychology, 10 (4), 375-392.
Liefooghe A.P.D. and Olafsson, R. (1999) ‘”Scientists” and “amateurs”:
Mapping the bullying domain.’ International Journal of Manpower, 20, 1&2,
39-49.
72
Littler, C.R., Bramle, T. & MacDonald, J . (1994) Organisational restructuring:
downsizing, delayering and managing change at work. Industrial Relations
Research Series, No. 15, AGPS, The Department of Industrial Relations,
Canberra, December.
Murphy, L.R. and Sauter, S.L. Work Organization Interventions: State of
Knowledge and Future Directions. Cincinnati (OH), National Institute for
Occupational Safety and Health.
Neuman, J .H. & Baron, R.A. (1996) Workplace violence and workplace
aggression: evidence concerning specific forms, potential causes, and
preferred targets. Journal of Management, 24, 391-419.
Quinlan, M. (1999) The implications of labour market restructuring in industrial
societies for occupational health and safety. Economic and Industrial
Democracy, 20, 427-460.
Rayner, C. (1999). From research to implementation: Finding leverage for
prevention. International Journal of Manpower, 20, 28-38.
Rayner, C., Hoel, H. & Cooper, C.L. (2002) Workplace Bullying: What We
Know, Who is to Blame, and What Can We Do? London: Taylor and Francis
Robinson, S. L. (1996) Trust and Breach of the Psychological Contract.
Administrative Science Quarterly, 41: 574-599.
Rousseau, D.M. (1995) Psychological Contracts in Organisations:
understanding written and unwritten agreements. Thousand Oaks: Sage.
Salin, D. (2004) Ways of explaining workplace bullying; a review of enabling,
motivating and precipitating structures and processes in the work environment.
Human Relations, 56.
Lewis, D. & Sheehan, M. (2003) Workplace bullying: theoretical and pratical
approaches to a management challenge. International Journal of Decision
Making, 4: 1-10.
Sparks, K. & Cooper, C.L. (1999) Occupational differences in the work-strain
relationship: Towards the use of situation-specific models. Journal of
Occupational and Organizational Psychology, 71, 219-229.
Spurgeon, A. (2003) Bullying from a risk management perspective. . In
S.Einarsen, H.Hoel, D.Zapf & C.L.Cooper (Eds.) Bullying and Emotional
Abuse in the Workplace. International perspectives in research and practice.
(327-338). London/New York: Taylor and Francis.
Stewart, M.B. and Swaffield, J .K. (1997) ‘Constraints on the desired hours of
work of British men. Economic Journal, 107, 520-535.
Turner, B.A. (1978) Man-made Disasters. London: Wykeham.
73
UNISON. (1997). UNISON members experience of bullying at work. London:
UNISON.
UNISON (2000) Police Staff Bullying Report (number 1777), London:
UNISON
Vartia, M. (1996) The sources of bullying: Psychological work environment
and organisational climate. European Journal of Work and Organizational
Psychology, 5, 203-214.
Wharton (1992) Risk mmanagement: basic concepts and general principles.
In J . Ansell and F. Wharton (eds.). Risk Analysis, Assessment and
Management, Chichester: J ohn Wiley.
Zapf, D., Einarsen, S., Hoel, H., & Vartia, M. (2003). Empirical findings on
bullying. In S. Einarsen, H. Hoel, D. Zapf, & C. L. Cooper (Eds.), Bullying and
Emotional Abuse in the Workplace. International perspectives in research and
practice (103-126). London: Taylor and Francis.
Zapf, D., Knorz, C., & Kulla, M. (1996). On the relationship between mobbing
factors, and job content, the social work environment and health outcomes.
European Journal of Work and Organizational Psychology, 5, 215-237.
74
16. Appendices
Appendix A: Advisory Board Membership
Baroness Anne Gibson Chair of Committee,
Member of the House of Lords.
Rob Allan Director of Human Resources,
West Hertfordshire NHS Trust.
Janet Asherson Head of Environment, Health and Safety Group,
Confederation of British Industry (CBI).
Chris Ball Independent Consultant / Amicus-MSF.
Jan Berry Chairman,
Police Federation of England and Wales.
Petra Cook Chartered Management Institute
Karen Charlesworth
Valery Davey Former Member of Parliament for Bristol West.
Angela Ishmael The Andrea Adams Trust
Brian Kazer Chief Executive,
British Occupational Health Research Foundation.
Diana Lamplugh The Suzy Lamplugh Trust
Barbara Lindsay MBE Senior Policy Adviser,
Department of Trade and Industry.
Tom Mellish Trades Union Congress
David Palferman Health and Safety Executive
Scott O’Brien
Dianah Worman Chartered Institute of Personnel and Development.
Keith Handley
Imogen Haslam
Helge Hoel The University of Manchester
Sabir Giga
75
Appendix B: The Bullying Risk Assessment Tool (BRAT)
The following items relate to your experience within your organisation. Please rate each item
by circling the number that best corresponds to your experiences / thoughts over the last 6
months.
1
Strongly agree
2
Agree
3
Slightly agree
4
Slightly disagree
5
Disagree
6
Strongly disagree
1. New staff are made to feel welcome when starting employment in the
organisation
1 2 3 4 5 6
2. Conflict in my work unit is common 1 2 3 4 5 6
3. I am clear about what is expected from me 1 2 3 4 5 6
4. This organisation does not value equal opportunity for everyone 1 2 3 4 5 6
5. I have confidence in my line managers abilities 1 2 3 4 5 6
6. Staff shortages are common in my unit 1 2 3 4 5 6
7. I enjoy working in the teams that I am involved with 1 2 3 4 5 6
8. I am not clear about how to carry out my job 1 2 3 4 5 6
9. Vacant positions are filled without delay within my unit 1 2 3 4 5 6
10. My line manager tries to control every single aspect of what is going on at work 1 2 3 4 5 6
11. The grading / rank structure in this organisation is transparent 1 2 3 4 5 6
12. I don’t get on with some of my colleagues 1 2 3 4 5 6
13. I have received sufficient training to carry out my job 1 2 3 4 5 6
14. My unit often makes use of temporary staff 1 2 3 4 5 6
15. My line manager values constructive criticism 1 2 3 4 5 6
16. People in this organisation are not rewarded properly 1 2 3 4 5 6
17. I find my colleagues to be co-operative 1 2 3 4 5 6
18. I face conflicting demands in my job 1 2 3 4 5 6
19. Cover for absent staff is provided immediately within my unit 1 2 3 4 5 6
20. My line manager exploits his / her position of power 1 2 3 4 5 6
21. I feel my contribution to the organisation is recognised 1 2 3 4 5 6
22. Different professional groups don’t work well together within my unit 1 2 3 4 5 6
23. My job description is clearly defined 1 2 3 4 5 6
24. I feel that there isn’t enough time in the day to complete my work 1 2 3 4 5 6
25. My line manager consults me before decisions affecting me are made 1 2 3 4 5 6
26. The organisations’ resources are not distributed fairly 1 2 3 4 5 6
27. My line manager is sensitive to how I feel 1 2 3 4 5 6
28. Existing work pressures make it difficult to take time off work 1 2 3 4 5 6
29. Work is shared equally among the people I work with 1 2 3 4 5 6
76
Factor structure of BRAT factor structure
1. ‘Organisational fairness’: 1, 11, 21, 4, 16, 26
2. ‘Team conflict’: 2, 12, 7, 17, 22, 29
3. ‘Role conflict’: 8, 3, 18, 13, 23
4. ‘Workload’: 6, 14, 24, 28, 9, 19
5. ‘Leadership’: 5, 10, 15, 25, 27, 20
77
Appendix C: Validation of the BRAT (results of regression analysis)
Table A1: BRAT factors as independent predictors of NAQ-R
Beta t Sig.
BRAT – Org. fairness 0.129 3.310 P<0.001
BRAT – Team conflict 0.202 5.259 P=001
BRAT – Role conflict 0.155 4.216 P<0.001
BRAT – Workload (6) 0.107 3.174 P=0.002
BRAT - Leadership 0.207 5.637 P<0.001
Rsq =34.2%
Table A2: BRAT factors as independent predictors of NAQ-R - Work-related
harassment
Beta t Sig.
BRAT – Org. fairness 0.125 3.428 p=0.001
BRAT – Team conflict 0.115 3.311 p=0.001
BRAT – Role conflict 0.241 7.067 p<0.001
BRAT – Workload (6) 0.188 5.928 p<0.001
BRAT - Leadership 0.288 6.566 P<0.001
Rsq=41.2%
Table A3: BRAT factors as independent predictors of NAQ-R – Personal harassment
Beta t Sig.
BRAT – Org. fairness 0.116 2.755 p=0.006
BRAT – Team conflict 0.250 5.990 p<0.001
BRAT – Role conflict 0.055 1.383 NS
BRAT – Workload (6) 0.019 0.521 NS
BRAT - Leadership 0.162 4.019 p<001
Rsq=22.3%
Table A4: BRAT factors as independent predictors of self-labelled bullying
Beta t Sig.
BRAT – Org. fairness 0.129 3.119 p=0.002
BRAT – Team conflict 0.315 7.736 p<0.001
BRAT – Role conflict -0.099 -2.536 p=0.011
BRAT – Workload (6) 0.032 0.876 NS
BRAT - Leadership 0.192 4.848 p<001
Rsq=23.9%
Table A5: BRAT factors as independent predictors of General Health Questionnaire
(GHQ-12)
Beta t Sig.
BRAT – Org. fairness 0.066 1.584 NS
BRAT – Team conflict 0.107 2.565 p=0.11
BRAT – Role conflict 0.314 7.996 p<0.001
BRAT – Workload (6) 0.053 1.469 NS
BRAT - Leadership 0.100 2.488 p=0.013
Rsq-24.4%
78
Appendix D: Policy guidelines
NB: Bullet points in bold considered essential, other points considered desirable.
Statement of intent and commitment
• The right to work in an environment free of harassment, bullying &
intimidation (– working towards an environment free of
harassment/bullying).
• The seriousness of the problem, e.g. potential that disciplinary
actions may be taken, including dismissal. Can be unlawful/criminal
offence
• Applies to all employees, managers, workers, individuals
subcontracted or seconded to work for the organisation.
• The responsibility of all employees to comply with the policy
• Managers responsible for implementation of policies
• No recrimination/victimization
• Focus on a standard of conduct
Definitions & examples of behaviour and conduct in breach of policy
• Provide definition
• Examples of behaviours in breach of policy
• Highlights perception of recipient of exposure to unwanted
behaviour/bullying refers to behaviour which is unwanted by the
recipient
• Focus on deed as oppose to intent of behaviour acknowledging that
harassment and bullying can be unintentional
• (reference to a ‘reasonable’ person’s judgment of what constitutes
bullying
• Acknowledges/makes a distinction between one-off negative acts as
oppose to repeated negative behaviour
Principles for a safe complaint system: Reassurance of fairness, non-
recrimination and confidentiality
• No attempt of recrimination of targets (the person who files the
complaint) will be tolerated
• Confidentiality should be offered to the complainant (target) as far is
possible for the progression of the case. Unconditional
confidentiality cannot be offered as it may compromise the
employer’s general ‘duty of care’
• Principles of fairness and non-recrimination extends to alleged
perpetrators
• A message that complaints will be taken seriously
• Malicious complaints is considered a disciplinary offence
• Fairness to be assured by means of management training and
consistency in application of rules and regulations (reducing the impact of
subjectivity)
79
How to complain, seeking advice
• Line-managers should under normal circumstances be the first line
of contact
• Where to complain if the perpetrator happens to be the line-manager,
e.g. dedicated HR person/representative
• Emphasising the right to be accompanied when filing a complaint
• Outline the status and role of advisors and how the may be
contacted
o to provide advice on the rights of targets and the alternative
actions available to targets
o to provide practical help, e.g. assisting in drafting letters,
assisting targets to meetings etc.
• Professional support/counseling to be offered to target (complainant) as
well as alleged perpetrator throughout process
Reporting and complaints procedures
• Clear ad unambiguous procedures for resolving issues and
complaints
• Making a distinction between and informal and an informal complaint
rout
• Informal complaint (involving as few people as possible and with
interviews and discussions as far as possible held in an informal
atmosphere)
o Use of informal complaint does not prevent later use of the
formal complaint procedure
o Independence of investigators as far as possible
o Granting leave/suspension (with pay) during investigation if
necessary
• Formal complain
o Clarify how and where (to whom) complaint should be made
o Swift response, stating the given time period within which a
response will be made, e.g. a week
o Emphasise that the target should feel free to be accompanied
by a person of their choice in interviews,, e.g. a colleague or a
shop steward
o Potential sanctions listed
Information about how the policy is monitored (examples of potential approaches)
• Register of complaints/incidents (and their outcomes) to be retained
by HR
• Regular review of policy and monitoring system
• Collect information on negative behaviour as part of exit-interviews
• Include questions on the effectiveness of policy as part of staff surveys
80
Appendix E: Part A Tool: Bullying Risk Assessment Tool
Objective Hazard Identification: To be completed by the Human Resources
Department
1. Unit identification code: _______________
2. Total number of employees in unit_______
3. What is the average annual absenteeism rate in the unit*?
_______days per employee.
4. What is the average annual absenteeism rate in the organisation?
_______days per employee.
5. Is the unit experiencing specific cases of long-term absenteeism due to non-physical
illness?
Yes ? No ?
6. Are any employees in the unit often absent for shorter periods of time (i.e. 1 day at a time)
without explanation?
Yes ? No ?
7. What is the current annual employee turnover rate (percentage) in the:
(a) unit _______%
(b) organisation _______%
8. What was the annual employee turnover rate (percentage) in the previous year in the:
(a) unit _______%
(b) organisation _______%
9. How many complaints/grievances have been filed by employees from the unit in the last
year?
________
10. To what extent have the following affected the unit during the past 6 months:
Not at all
1
Some
2
Considerably
3
Extensively
4
Organisational / technological change 1 2 3 4
Financial budget cutbacks 1 2 3 4
Redundancies 1 2 3 4
Staff shortages / unfilled posts 1 2 3 4
Change of manager / supervisor 1 2 3 4
*Note: Unit refers to the level of analysis and will vary in each organisation (For example a ward in an
NHS Trust).
81
Appendix F: Demographic variations in the experience of bullying to bullying
Table A6: Gender differences in exposure to bullying
Male (%) Female (%) Sig.
Civil Service 9.2 15.3 p=.075
NHS 1 10.3 9.7 p=.499
NHS 2 12.5 17.1 p=.230
NHS 3 25.0 14.0 p=.152
Police 11.9 14.3 p=.313
Table A7: Age differences in exposure to bullying
Table A8: Contractual arrangement differences in exposure to bullying
Table A9: Hours of work differences in exposure to bullying
16-24
(%)
25-34
(%)
35-44
(%)
45-54
(%)
55-70
(%)
Sig.
Civil Service 13.3 8.6 14.9 15 12.1 p=.636
NHS 1 0 9.5 9.1 7.9 13.3 p=.923
NHS 2 0 22.7 13.3 13.2 25 p=.811
NHS 3 0 24.2 14.5 14.9 11.5 p=.936
Police 0 10.0 16.2 11.1 18.7 p=.789
Permanent
(%)
Fixed
period
(%)
Seconded
(%)
Other
(%)
Sig.
Civil Service 12.6 0 0 n/a p=1.000
NHS 1 11.0 0 0 0 p=1.000
NHS 2 15.9 25.0 n/a n/a p=.767
NHS 3 14.9 40.0 0 100 p=.002
Police 13.5 9.1 n/a 0 p=.997
Full-time (%) Part-time (%) Sig.
Civil Service 12.7 10.0 p=.633
NHS 1 10.9 6.8 p=.794
NHS 2 20.2 8.3 p=.382
NHS 3 20.6 7.4 p=.121
Police 12.3 18.2 p=.232
82
Table A10: Ethnic differences in exposure to bullying
Table A11: Organisational level of target
No
spvsr.
resp. (%)
Spvsr.
(%)
Middle
Mgmnt.
(%)
Snr.
Mgmnt.
(%)
Other
(%
Sig.
Civil Service 13.3 10.3 9.0 22.7 33.3 p=.243
NHS 1 14.7 8.9 6.2 0 5.3 p=.264
NHS 2 15.6 11.8 6.7 0 38.5 p=.433
NHS 3 13.5 15.4 24.2 33.3 12.5 p=.711
Police 11.9 17.4 17.2 0 14.3 p=.951
Table A12: Duration of bullying experience: breakdown within organisations
White (%) Non-white (%) Sig.
Civil Service 10.4 22.7 p=.010
NHS 1 9.8 10.5 p=.693
NHS 2 14.5 30.8 p=.317
NHS 3 12.8 34.4 p=.035
Police 12.2 33.3 p=.021
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Civil Service 32.3 22.6 22.6 22.6
NHS 1 5.3 15.8 21.1 57.9
NHS 2 9.5 23.8 38.1 28.6
NHS 3 23.5 23.5 29.4 23.5
Police 20.0 20.0 20.0 40.0
83
Table A13: Negative Acts: Breakdown within organisations
*(R)egularly or (O)ccasionally experiences
*
Someone withholding
information which affects
your performance (%)
Having your opinions and
views ignored (%)
Being exposed to
an unmanageable
workload (%)
O
45.9 37.7 34.7 Civil Service
R 8.4 7.3 11.9
O
47.0 37.0 36.2 NHS 1
R 11.9 8.7 14.1
O
48.6 45.3 42.1 NHS 2
R 7.9 8.7 10.7
O
44.8 46.2 30.3 NHS 3
R 10.5 8.1 17.5
O
49.0 38.0 29.3 Police
R 10.8 8.7 13.6
84
Appendix G: Pre- and post-intervention data comparisons
Table A14: Negative behaviour (no significant changes in terms of high
ranking negative behaviours)
*(R)egularly or (O)ccasionally experiences
Table A15: Duration of bullying experience
* Pre-
intervention
(%)
Post-
intervention
(%)
O 47.0 48.9 Someone withholding information which affects your
performance
R 10.0 9.4
O 40.4 40.6 Having your opinions and views ignored
R 8.2 7.3
O 33.8 34.7 Being exposed to an unmanageable workload
R 13.7 14.2
O 31.0 30.9 Being ordered to do work below your competence
R 9.4 11.0
O 29.2 29.9 Being given tasks with unreasonable or impossible
targets or deadline
R 8.7 10.3
O 24.5 25.1 Having key areas of responsibility removed or replaced
with more trivial or unpleasant tasks
R 6.9 6.6
O 22.5 22.5 Spreading gossip
R 4.8 5.2
O 22.4 23.2 Being shouted at or being the target of spontaneous
anger R 4.7 3.7
O 22.3 18.9 Being humiliated or ridiculed in connection with you
work
R 4.6 4.0
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Pre-intervention 20.0 21.5 25.9 32.6
Post-intervention 25.2 21.8 22.7 30.3
85
Table A16: Organisational breakdown of bullying duration
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Civil Service (T0) 32.3 22.6 22.6 22.6
Civil Service (T1) 24.1 34.5 13.8 27.6
NHS 1 (T0) 5.3 15.8 21.1 57.9
NHS 1 (T1) 29.4 11.8 35.3 23.5
NHS 2 (T0) 9.5 23.8 38.1 28.6
NHS 2 (T1) 41.7 12.5 20.8 25.0
NHS 3 (T0) 23.5 23.5 29.4 23.5
NHS 3 (T1) 17.2 20.7 27.6 34.5
Police (T0) 20.0 20.0 20.0 40.0
Police (T1) 15.0 25.0 20.0 40.0
doc_888700757.pdf
Workplace bullying has been increasingly acknowledged as a major concern to British employers and their workers, with implications for individuals as well as organisations.
Destructive Interpersonal Conflict
in the Workplace:
The Effectiveness of Management
Interventions
Funded by the
British Occupational Health Research Foundation (BOHRF)
Helge Hoel & Sabir I. Giga
With contributions from Brian Faragher
Manchester Business School
The University of Manchester
February, 2006.
©Helge Hoel & Sabir I. Giga, 2006.
2
Contents
Acknowledgements.................................................................................................... 4
Executive Summary................................................................................................... 5
1.1 Introduction.......................................................................................................... 7
1.2 Background.......................................................................................................... 8
1.3 Key findings from previous intervention research................................................ 9
2. Research method................................................................................................. 10
2.1 Focus groups.................................................................................................. 11
2.2 Questionnaire survey..................................................................................... 11
2.3 Interventions................................................................................................... 12
2.4 Research design............................................................................................ 12
2.5 Local steering committees.............................................................................. 13
3. Pre-intervention focus groups.............................................................................. 14
4. Development of the Bullying Risk Assessment Tool (BRAT)............................... 15
4.1 Factor structure.............................................................................................. 17
4.1.1 ‘Organisational fairness’........................................................................... 17
4.1.2 ‘Team conflict’ .......................................................................................... 17
4.1.3 ‘Role conflict’............................................................................................ 18
4.1.4 ‘Workload’ ................................................................................................ 18
4.1.5 ‘Leadership’ ............................................................................................. 19
4.2 Factor reliability.............................................................................................. 19
4.3 Validation of the BRAT................................................................................... 20
4.4 Application of the BRAT ................................................................................. 21
5. Identification and development of interventions................................................... 21
5.1 Policy communication..................................................................................... 22
5.2 Stress management training programme....................................................... 22
5.3 Negative behaviour awareness training programme...................................... 23
5.4 Implementation of interventions...................................................................... 24
6. Questionnaire survey participants and sample background................................. 25
6.1 Sample - study participants............................................................................ 25
6.2 Sample background....................................................................................... 26
6.2.1 Gender..................................................................................................... 26
6.2.2 Age .......................................................................................................... 27
6.2.3 Ethnicity ................................................................................................... 28
6.2.4 Contracted hours of work......................................................................... 28
6.2.5 Employment contract ............................................................................... 29
6.2.6 Job type ................................................................................................... 29
6.2.7 Supervisory responsibilities ..................................................................... 29
6.2.8 Length of time with organisation .............................................................. 30
7. Objective data...................................................................................................... 30
8. Experiences of bullying in the workplace: baseline measures ............................. 31
8.1 Experience of bullying.................................................................................... 31
8.2 Witnesses of bullying...................................................................................... 33
8.3 Past bullying................................................................................................... 33
8.4 Targets of bullying.......................................................................................... 34
8.4.1 Gender..................................................................................................... 34
8.4.2 Age .......................................................................................................... 34
8.4.3 Full-time versus part-time ........................................................................ 35
8.4.4 Ethnicity ................................................................................................... 35
3
8.5 Organisational level of target.......................................................................... 36
8.6 Perpetrators of bullying .................................................................................. 36
8.7 How many were bullied? ................................................................................ 37
8.8 Duration of the bullying experience................................................................ 37
8.9 Experiences of negative behaviours............................................................... 37
8.10 Outcomes of bullying.................................................................................... 38
9. Experiences of bullying in the workplace pre and post intervention
comparisons .................................................................................................... 39
9.1 Experiences of bullying.................................................................................. 39
9.2 Perpetrators.................................................................................................... 40
9.3 How many were bullied? ................................................................................ 41
9.4 Psychological contract.................................................................................... 42
10. Efficacy of intervention: comparing baseline and post-intervention
measures for different interventions across organisations............................... 43
10.1 Overall results .............................................................................................. 43
10.2 Bullying......................................................................................................... 44
10.3 Witnessing bullying....................................................................................... 45
10.4 Negative behaviours..................................................................................... 46
10.5 The psychological contract........................................................................... 47
10.6 Mental health................................................................................................ 49
10.7 Organisational outcomes, absenteeism and intention to leave..................... 50
10.8 Satisfaction with the atmosphere at work..................................................... 51
10.9 The effectiveness of intervention: assessing ‘objective’ measures............... 53
10.10 Intervention and levels of absenteeism...................................................... 53
10.11 Other ‘objective’ measures......................................................................... 54
10.12 Total assessment of effectiveness of interventions .................................... 54
11. Intervention feedback from the trainer and participants ..................................... 56
11.1 The trainer’s perspective.............................................................................. 56
11.2 Feedback from participants.......................................................................... 57
12. Post Intervention focus groups........................................................................... 58
13. Discussion ......................................................................................................... 63
14. Conclusions ....................................................................................................... 68
15. References......................................................................................................... 70
16. Appendices........................................................................................................ 74
Appendix A: Advisory Board Membership............................................................ 74
Appendix B: The Bullying Risk Assessment Tool (BRAT).................................... 75
Appendix C: Validation of the BRAT (results of regression analysis)................... 77
Appendix D: Policy guidelines.............................................................................. 78
Appendix E: Part A Tool: Bullying Risk Assessment Tool .................................... 80
Appendix F: Demographic variations in the experience of bullying to bullying..... 81
Appendix G: Pre- and post-intervention data comparisons .................................. 84
4
Acknowledgements
We would like to thank the British Occupational Health Research Foundation
(BOHRF) for funding this research and in particular, Brian Kazer, the Chief
Executive of BOHRF for his unremitting support throughout the duration of
this project. Our unreserved gratitude to members of our Advisory Board
(Appendix A), not least the Chair – Baroness Gibson of Market Rasen, for
their guidance and constructive critique during various stages of the research.
We would also like to acknowledge the support of the five organisations who
took part in this study for not only agreeing to participate but also for
demonstrating enduring commitment, and to Rob Allan and the West
Hertfordshire NHS Trust for allowing us to pilot our interventions within their
organisation. Our appreciation also to Professor Cary Cooper for his support
in initiating this study as well as his continued interest in our work. Finally, we
would like to thank our colleague Dr Brian Faragher for his invaluable advice
on statistical analysis and our partners Neil and Tisha for their everlasting
patience.
5
Executive Summary
Workplace bullying has been increasingly acknowledged as a major concern
to British employers and their workers, with implications for individuals as well
as organisations. Whilst most studies to date have focused on the prevalence
of the phenomena and its consequences, very few have looked at remedial
action. To address this imbalance, an intervention study was carried out in
five British public sector organisations.
Sponsored by the British Occupational Health Research Foundation (BOHRF)
the aim of the study was two-fold: 1) to devise and test the appropriateness as
well as effectiveness of a risk assessment tool, and 2) to develop, implement
and evaluate three different bullying intervention programmes. These
programmes focussed on training in three different areas: policy
communication, stress management and negative behaviour awareness.
With regard to study method, a randomised control design was deemed
necessary to enable the researchers to make any degree of inference with
regard to causal relationships, with the same interventions carried out in
various combinations in all five organisations. Pre and post intervention data
were obtained by means of a questionnaire comprising of a variety of
instruments to measure negative behaviour and experiences and
consequences of bullying. This was supplemented with data from post-
intervention focus groups and information from the trainer as well as training
participants. In line with the study’s aims, the questionnaire survey also
contained an instrument to measure potential risk-factors of bullying – the
Bullying Risk Assessment Tool (BRAT). Informed by focus group findings as
well as expert opinion and a literature review, the refined version of the BRAT
has a factor structure made up of five factors: organisational fairness, team-
conflict, role-conflict, workload and leadership. A validation study confirmed
the validity of the instrument with all five factors emerging as predictors of
negative behaviour and self-labelled bullying.
Pre-intervention base-line measures were based on a sample of 1,041
respondents, a response-rate of 41.5%. The pre-intervention questionnaire
study confirmed that bullying is a problem in public sector organisations in the
UK with 13.6% of respondents stating that they have been bullied within the
last six months, compared with a national average of 10.6% obtained from a
previous BOHRF sponsored nationwide study (Hoel & Cooper, 2000).The
level of bullying in the current study varied between organisations, with the
lowest at 10.8% and highest at 16.6%. An analysis of the data highlights a
strong association between bullying and mental health and intention to quit,
thus confirming that bullying appears to have individual as well as
organisational implications. A total of 61% of perpetrators were reported to be
managers or supervisors whilst colleagues were considered to be the culprits
in 42% of incidents.
The main stress management and negative behaviour awareness
interventions were delivered by a professional trainer as three-hour training
6
sessions, whilst the policy communication was carried out over a thirty-minute
session. The training programme was implemented in all five organisations
over a six month period and involved approximately 150 participants in total.
Feedback contained in 193 training evaluation forms (participants on the full-
day training programme were required to evaluate the morning and afternoon
sessions separately), suggests that the training was well received, with a
substantial number of participants considering the sessions very interesting as
well as relevant. With regard to the trainer’s own end-of-training session
comments, it was emphasised that sufficient time was needed in order to
establish a climate in which constructive group dynamics and learning could
take place.
Post-intervention measures were obtained approximately six months after the
training had taken place by means of distributing the same questionnaire a
second time within the same units in all five participating organisations. This
made it possible to compare post and pre-intervention data for a large number
of variables relevant to the issue of bullying and negative behaviour at work.
Important improvement in the desired direction as measured by a variety of
variables did occur for 45% of the experiment groups. For three of the
experiment groups scores on all relevant variables measured were in the
desired direction. In two of these cases the experiment group had received a
combination of all three training programmes or workshops whilst the third
group had received the ‘negative awareness training’ in addition to policy
communication. The fact that it is difficult to make any firm conclusions with
regard to efficacy of a specific intervention is not surprising and is in line with
much organisational research. Threat of redundancy and in some cases a
strong unwillingness to engage with the training might also help explain why
the desired development did not take place in all experiment groups. Other
factors that may have possibly influenced the results include: increased
awareness and greater expectations which in turn lead to dissatisfaction
resulting from the training process; the right people, in this case managers
and supervisors, not taking part in the training in sufficient numbers; overall
‘critical mass’ not being achieved due to too few people being trained in order
to have a significant impact upon behaviour; and, finally, the time between
interventions and post-intervention measurement might have been too short
for significant effects to have occurred.
In conclusion, this research has resulted in the successful completion of the
first academic anti-bullying intervention study, comparing the effectiveness of
interventions across different organisational contexts and involving the
implementation of a complex design in order to apply scientific rigor. A risk
assessment tool focussing on issues relating to negative behaviour and
bullying has been developed and its properties validated. Although the study
was unable to establish beyond doubt the efficacy of a particular intervention,
there is evidence to suggest that theoretically sound, well planned and aptly
delivered interventions can make a difference, particularly when sufficient time
is allocated and the proportion of staff being trained is significant enough to
have an impact upon behaviour.
7
1.1 Introduction
This report presents the results of the very first academic study evaluating the
effectiveness of a variety of interventions to tackle workplace bullying and
highlights the opportunities organisations have in preventing and managing
this problem. Following on from the previous collaboration between the
University of Manchester Institute of Science and Technology (UMIST) and
the British Occupational Health Research Foundation (BOHRF) which
established the prevalence of workplace bullying (the first such study of its
kind in Great Britain), and having identified a number of sources associated
with negative behaviour in the workplace, the overall aim of this project was to
develop, implement and evaluate the effectiveness of organisational
interventions to prevent and manage destructive interpersonal conflict.
Based on the successful application of risk assessment to physical hazards in
the workplace, recent developments have seen attempts to apply such
techniques to the area of psychosocial hazards. In addition to identifying
effective intervention strategies in the present study, we also aimed to take
advantage of these theoretical gains to develop a risk assessment tool which
could assist managers and their organisations in managing and controlling
antecedents of destructive behaviour and bullying in the workplace.
We envisage that our findings will help inform academic and organisational
practice in the development of effective intervention and risk management
strategies to deal with interpersonal conflict in the workplace and as such
make an important contribution to the development of this relatively new field
of research and practice.
Structure of the report:
After providing a brief background to the study, the methodology is outlined in
section 2. The next section reports on the pre-intervention focus groups
carried out in order to inform the interventions as well assist in the
development of a risk assessment tool. Section 4 focuses on the development
and validation of the Bullying Risk Assessment Tool (BRAT) and section 5
examines the development and implementation of various interventions
identified to combat bullying and destructive behaviour. The study sample and
sampling procedure is then outlined in section 6. The findings of the two
surveys, baseline and post-intervention data, with measurements of a number
of variables associated with bullying and destructive behaviour are
investigated in sections 8 and 9. With the aim of establishing potential efficacy
of particular interventions or combination of interventions, a number of
analyses of the data were carried out. The results of these analyses are
critically examined in section 10. This is followed by a presentation of the
qualitative data based on feedback from study participants and other
observers, including comments from the trainer. Finally, in the closing
sections we discuss the results and highlight the main learning points for
future intervention studies before reviewing our key findings, and delivering
our closing opinions and conclusion.
8
1.2 Background
A number of studies, in the UK and abroad, have recently identified
destructive conflict and bullying at work as an occupational hazard of
significant magnitude (Zapf et al., 2003). However, it was not until the first
BOHRF funded study (Hoel & Cooper, 2000) that it was possible to establish
the prevalence of the problem in the UK with a high degree of certainty taking
into account the size of the sample and the research methodology applied.
The data from this survey suggests that approximately 10% of the UK
workforce consider themselves at any one time to be bullied. For a majority of
these targets of bullying, the experience is reported to be occasional rather
than regular. If the time frame is changed to include experience within the last
five years, approximately 25% of respondents report themselves as having
been bullied. Furthermore, when observers or witnesses of bullying, who are
not themselves targeted directly, are included, nearly half the respondents
report having had some first-hand experience of the problem.
The nationwide study sponsored by BOHRF also confirms previous findings
which suggest that being exposed to destructive conflict and bullying may
have implications for individuals (see Einarsen et al., 2003), not only with
regard to their mental and physical health, but also with respect to their
organisational commitment and satisfaction. Industry-specific analysis also
suggests that the same behaviours may be present (though to a varying
degree) but are interpreted in various ways and have a different impact within
different organisational settings (Hoel, et al., 2004). This result is in line with
recent findings in stress research, which emphasise that the relationship
between work-related stressors and strain (i.e. ill health) should be
understood in a local or situation-specific context (Giga et al., 2003; Sparks &
Cooper, 1999).
With regard to organisational implications, the findings from the nationwide
study (and studies carried out by Charlotte Rayner for Unison, e.g. Rayner,
1999) suggest a relationship between bullying and intention to leave the
organisation, with targets of bullying substantially more likely than non-targets
to leave. Furthermore, for the first time a large-scale study has identified a
similar, albeit weaker, association between exposure to bullying, on the one
hand, and absenteeism and productivity, on the other, suggesting that
employees who are targets of bullying may on average have seven days more
sick-leave than those who were neither bullied nor had witnessed bullying.
Moreover, the negative effects of bullying appear to extend beyond those who
report being bullied at present to include those who were bullied in the past as
well as those who have witnessed bullying taking place within the same time-
frame, but who have not themselves been directly targeted (Hoel, Faragher &
Cooper, 2003). It could, therefore be argued that it should be in the self-
interest of organisations to take this issue seriously and attempt to manage it
effectively.
Whilst the antecedents of bullying will vary between organisations, most
organisations at the beginning of the new millennium are still in the midst of or
struggling with the effects of significant change processes. External pressures
9
have forced organisations to restructure, downsize and become leaner (Littler
et al, 1994). This is true for the private as well as public sector. As a result,
employees at all levels of the organisation, in the private as well as in the
public sector find themselves in a position of increasing workloads, often in a
climate of uncertainty with regard to their current and future employment
(Stewart & Swaffield, 1997). In this situation managers are frequently required
to put in long and even excessive working hours whilst accountability and
responsibility for the human resources function is often devolved to the level
of the line manager (J ames, 1993). Faced with this situation, it is not
surprising to find that managers allegedly frequently make use of authoritarian
and even abusive behaviours in order to carry out their work (Sheenan, 1999).
In light of such pressures it might come as no surprise that a common finding
from research into bullying in the UK, that in three out of four incidents the
alleged perpetrator appears to be someone who has supervisory or
managerial responsibilities (Unison, 1997; Hoel, Faragher & Cooper, 2001).
Most studies of workplace bullying have so far focused on establishing the
prevalence of the phenomena and its consequences By contrast, despite
substantial effort by practitioners in addressing the issues in the workplace
(e.g. Rayner, Hoel & Cooper, 2002; Hubert, 2003) little evidence has emerged
with regard to effective interventions. Furthermore, references to
organisational interventions within the emerging bullying literature tend to rely
on propositions inferred from related areas such as workplace violence,
sexual harassment and stress (Hoel et al., 1999). According to the literature
on interventions it is noteworthy that the evidence with regard to efficacy, in
particular with respect to psychosocial issues, is sparse, with a number of
observers highlighting the need for further attention to such issues (e.g.
Kompier et al., 1998; Cox et al., 2000; Murphy & Sauter, 2003). Thus, to our
knowledge there are no studies that have systematically attempted to assess
the efficacy of various forms of interventions aimed at reducing and
minimizing negative behaviour and bullying at work. Moreover, the fact that
incidence studies of workplace bullying in the UK consistently suggest that the
majority of perpetrators are to be found within supervisory and managerial
ranks suggests that behaviour of managers should be a focal point of any
intervention study. Therefore, such an approach would also seem to be the
most appropriate one from a cost-benefit perspective where resources are
limited.
1.3 Key findings from previous intervention research
By means of a review of the intervention literature (e.g. Cooper & Cartwright,
1997; Kompier et al., 1998; Cox et al., 2000; Giga et al., 2003; Murphy &
Sauter, 2003) we identified, and observed as far as possible in the present
study, a number of issues which were considered crucial in order to achieve a
successful outcome:
Most interventions studies focus on either the individual (predominantly)
or the organisation. Very few studies combine measures aimed at both
the individual and the organisation.
10
For any intervention to become successful it must be based on sound
theory. Moreover, this understanding cannot be the sole domain of
researchers but must be shared with the organisation and its
management.
Without commitment and support from senior echelons within
management, including provision of necessary resources and a clear
understanding of the theoretical underpinning of the interventions any
initiatives are likely to fail. The fact that the presence of bullying is
considered as ultimately resting on managerial consent also further
strengthens the case for active management involvement.
Current research methods are insufficient and progress in this area is
necessary in order to establish efficacy of interventions. In particular, it
is crucial to make use of more objective measures (data which can be
validated by others).
Any intervention needs to be contextualised and tailored to the
requirements of the organisations by means of employee participation.
More scientific rigor in intervention studies including obtaining accurate
baseline measures is needed. In particular, in order to be able to take
into consideration cause and effect, a random control design (RCT), is
recommended.
Whilst as far as possible observing principles for sound scientific
design, realism with respect to what is possible within the context of
organisations which are continuously experiencing change and where
participants are unlikely to remain as passive study objects.
2. Research method
The aims of this research were to:
devise and test the appropriateness as well as the effectiveness of a
risk assessment tool
develop, implement and evaluate three different bullying intervention
programmes
The study was carried out in five host organisations: These include:
1. Civil Service Department (London and North England)
2. NHS Mental Health Trust (South England)
3. Acute NHS Trust (North England)
4. Acute NHS Trust (South England)
5. Police Force (South England)
Although all five organisations were from the public sector, with the exception
of the two acute NHS trusts, they are all very different and therefore the
generalisability and transferability of our findings should not be affected.
In order to accomplish the aims of this research, a mixture of quantitative and
qualitative methods were applied. These are outlined below.
11
2.1 Focus groups
A focus group methodology was used at two different stages of the project.
The purpose for conducting focus groups during the initial stages of the
project within participating organisations included:
1. Contribution to the development of context-specific policies;
2. Contribution to the development of a risk-assessment tool; and
3. Ensuring that the intervention programmes were developed and carried
out with reference to the local context and to address issues identified
as important by participants.
Focus groups are seen as useful instruments due to their interactive qualities
and their ability to explore perceptions of individuals, and, in particular, the
subjective meaning individuals assign to events (Liefooghe & Olafsson, 1998).
Group interaction also has the potential to provide a range of views on the
issues under consideration (Kitzinger, 1994), a fact which is considered
invaluable when individual and context-specific meaning is being explored.
Bellenger et al. (1976) argue that when a number of people collectively
explore ideas by providing others with insight into their views and way of
thinking, the outcome of the interaction is more than the sum of its parts. The
‘snowball’ effect, in which a random comment may bring about a ‘chain-
reaction’ of responses, accounts for part of this synergetic effect. As
interaction was at the heart of the study, the primary role of the focus group
facilitator was to ensure that the discussions took their course. No attempt
was made to interfere in natural group processes.
In total, 55 focus groups were carried out between 01/10/2003 and 26/02/04
involving a total of 272 people. In most cases, approximately 7% of staff from
each unit were involved in this part of the study.
A second stage of focus groups were scheduled six months after the
interventions for the purpose of discussing changes in behaviour and in the
work environment following implementation of the interventions. Altogether
eight post-intervention focus groups were carried out in three organisations
2.2 Questionnaire survey
In order to evaluate the efficacy of our intervention programmes and the
validity of the risk-assessment tool, details of which are given in subsequent
sections, we developed a comprehensive questionnaire comprising of
questions related to experiences of bullying and negative behaviour, and
potential risk-factors of bullying. The questionnaire also contained measures
on mental health, sickness-absence, intention to quit, self-rated productivity,
job-satisfaction, individuals’ ‘psychological contract’ and a number of
demographic variables collected to make it possible to compare the
experience of different groups.
12
Some of the above sections comprised of existing validated instruments whilst
others such as the Bullying Risk Assessment Tool (BRAT) needed
development by the researchers (see section 4). The 22-item NAQ-R
(Einarsen & Hoel, 2001) was used to measure experiences of negative
behaviour and bullying. To assess the state of employee psychological
contracts we used Robinson’s (1996) 7-item measure of fulfilled employer
obligations and 2-item measure of met employment expectations. Finally,
mental health was measured using the 12-item General Health Questionnaire
(Goldberg & Williams, 1988).
As one of the main aims of this study was to devise and test the
appropriateness as well as the effectiveness of a risk assessment tool, we
specifically set out to develop a bespoke instrument for inclusion in our pre-
intervention questionnaire survey and took the opportunity of the post-
intervention survey to refine it further.
In order to identify any changes resulting from our interventions , the
questionnaire was distribute to the same pool of people twice - prior to the
delivery of the interventions and six months after the interventions.
2.3 Interventions
As indicated in the introduction, the intervention literature, not least with
respect to psychosocial hazards, advocates a context specific or local
approach responding to local needs and based on broad employee
involvement (e.g. Kompier et al., 1998; Giga et al., 2003). However, since our
main aim was to compare effectiveness of interventions across different
organisational settings, we concluded that it would not be feasible to compare
five completely different approaches from a scientific perspective. Such an
approach would also mitigate against generalisability as well as transferability
of findings to other contexts. However, although the interventions were based
on a common core approach, the local context was taken into consideration
and the programme was tailor-made to meet the need and reality of the
individual organisations by means of examples given and tasks set. On the
basis of a review of the literature and discussions with experts in the field we
identified the following three interventions: policy communication, stress
management training and negative behaviour awareness training. Each of
these interventions and their rational are outlined in section 5.
2.4 Research design
In order to avoid possible problems associated with research methodology,
such as selection bias and cross-contamination of data due to communication
between participants taking part in different interventions, the intervention
programme was implemented in similar or parallel departments or units by
means of a clustered randomised selection process. In each organisation five
groups were selected and assigned one of the following roles:
13
Figure 1: Research design
Group (1)
(Control grp)
Group (2) Group (3) Group (4) Group (5)
Policy
Communication
Policy
Communication
Policy
Communication
Policy
Communication
Stress
Management
Training
Negative
Behaviour
Awareness
Training
Stress
Management
Training AND
Negative
Behaviour
Awareness
Training
A sample of approximately 20-25 managers (or other employees) from each
group were invited to attend relevant training sessions. Group (1) served as a
control group and did not take part in any intervention, group (2) were invited
to a 30-minute policy communication session, group (3) were involved in a 3-
hour stress management training programme in addition to the policy
communication, group (4) were invited to a 3-hour negative behaviour training
session in addition to the policy communication session, and group (5) were
involved in a full day training session covering policy communication, stress
management and negative behaviour awareness. Although the researchers
ideally would have liked to increase the scope or length of training, initial
discussion with participating organisation suggested that any further time-
commitment was not feasible.
2.5 Local steering committees
From the start we encouraged the formation of local steering committees. The
overall aim of steering committees was to assist researchers in developing
and implementing the research programme locally. It was suggested that such
committees have a strong management presence, with representation of
occupational health practitioners and employees (unions/staff associations
where appropriate).
To ensure management commitment to the process we also required that
each participating organisation made a financial contribution to the study.
Furthermore, from the outset we emphasised that successful implementation
largely rested on continuous management support and warned against
potential impact of internal pressures, organisational change and influence of
competing initiatives.
In most organisations, a nominee from human resources served as the main
point of contact for the research team as well as a facilitator for the steering
group. Active support from senior management was encouraged from the very
early stages of the project.
14
3. Pre-intervention focus groups
The aims of the focus groups were to assist in the development of the risk
assessment tool as well as inform the development of specific interventions.
Apart from recording each focus group, participants were provided with a list
of negative behaviours taken from the revised Negative Acts Questionnaire
(NAQ-R)(Einarsen & Hoel, 2001), which is a validated instrument in its own
right and used extensively in bullying research. The one-page form also
requested simple demographic information and gave participants the
opportunity to list other negative behaviours specific to their organisation.
Information from both the interview transcriptions and feedback from
participants have been used in developing our risk assessment tool.
In total, 55 focus groups were scheduled between 01/10/2003 and 26/02/04
involving a total of 272 people. We piloted the exercise during the first two
sessions which involved 13 people from the Civil Service. The remaining
groups were scheduled as follows:
7 further groups within the Civil Service Department’s London sites
10 groups within the Acute NHS Trust (North)
9 groups within the Acute NHT Trust (South)
17 groups within the NHS Mental Health Trust
10 groups within the Police Force
A majority of the focus groups were carried out according to the protocol.
However, this process has not been without its problems, with some focus
groups cancelled due to organisations not being able to recruit sufficient
participants to make the groups viable. This was not least the case when
attempting to organise separate sessions for certain groups such as ethnic
minority staff. Whilst a majority of focus groups consisted of 3-6 people,
scheduling difficulties were also common, demanding maximum flexibility on
behalf of the facilitator. In one rather extreme example, due to
misunderstandings not less than 21 people attended the same session. In this
case the problem was resolved by turning what was meant to be a focus
group into a fact-finding group-meeting, obviously different from the nature of
the focus groups but still able to provide valuable input to the study. The
reasons for the low participation in some sessions differ from one organisation
to another. However, in general this was not due to a lack of interest on behalf
of prospective participants but instead difficulties in getting time off work from
what were considered already understaffed departments. It was particularly
difficult to engage employees who were based in units that were located some
distance apart. Although facilities have not been ideal for conducting focus
groups at these remote sites, we have in many cases insisted that some
sessions be held there in order not to exclude certain groups of employees.
The focus groups centred on two issues: assessing which negative behaviour
participant considered particularly difficult to handle or deal with and; what
factors, situations or antecedents contributed to negative behaviour and
bullying.
15
As suggested in the figure (2) below, results from focus groups highlighted
behaviour involving some degree of personal attack as most difficult to deal
with across organisational setting, with behaviours such as ‘being humiliated’
(item 2), ‘being ignored or excluded’ (item 6), ‘hints from others to quit your
job’ (item 10), and ‘personal allegations’ (item 17) scoring highly (3 refers to
very difficult to deal with and 0 refers to not at all difficult to deal with).
However, discrepancies exist both within and between organisations. For
example, certain negative behaviours such as ‘being ordered to work below
their level of competence’ (item 3) are seen as normal, unproblematic and for
some even welcomed at times, whereas as for others it is much more
problematical.
Figure 2: Difficult to deal with behaviours breakdown for each organisation
Focus group participants also identified behaviours that were not listed on the
NAQ(R) and which they found difficult to deal with. These were taken into
consideration in connection with development of the intervention programmes.
Other findings from the focus groups are highlighted in the next section.
4. Development of the Bullying Risk Assessment Tool (BRAT)
Aim: To devise and test the appropriateness as well as the effectiveness of a
risk assessment tool developed on the basis of the current study.
Risk assessment is a commonly used approach throughout industry to identify,
evaluate and control potential harmful effects of physical hazards. However, a
number of observers have advocated that a risk assessment strategy should
also be applied to psychosocial hazards such as occupational stress (e.g. Cox
et el., 2000) and bullying (Spurgeon, 2003). Central to such approaches is the
distinction between a hazard (something that has the potential to cause harm)
and risk (the probability that such a harm will occur). According to Spurgeon
risk-assessment should be part of a risk-management approach which would
incorporate the following elements or steps: 1) hazard identification, 2) risk
reduction and 3) control of risk. It is important to note that such an approach is
focused on reduction and control as opposed to elimination of risk because
many workplace hazards might not be altogether eradicated as they often
0.00
0.50
1.00
1.50
2.00
2.50
3.00
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Series1
Series2
Series3
Series4
Series5
16
represent part of normal activities and interactions in the workplace. In this
respect, the risk assessment process is focused on ‘estimating the probability
and the size of possible outcomes, and then evaluating the alternative
courses of action’ (Wharton, 1992 cited in Clarke & Cooper, 2004). As risk is
considered to be the product of the frequency and the consequences of
exposure to a hazard, it infers that hazards are likely to have severe
consequences.
Bullying can also be compared to what Turner (1978) refers to as ‘slow
accidents’, as the likelihood of bullying actually occurring may not always be
clear due to the uncertain effects of hazards (Clarke & Cooper, 2004). It is
also acknowledged that the intangible nature of psychosocial hazards, and
their subjective nature which is open to interpretation by individuals,
undermines the potential to measure or estimate the risk on a calibrated scale
as is the case to a far larger degree for physical hazards. Finally, as
highlighted in the stress literature, to be effective any risk assessment tool
needs to be context specific in the sense that it reflects the hazards within the
environment to which it is being applied (Cox et al, 2000; Giga, Cooper &
Faragher, 2003) and its success would, therefore, to a large extent be
contingent upon the degree of employee involvement and management
commitment to the process.
These factors need to be taken into consideration when devising a risk
assessment tool for it to be effectively be applied to workplace bullying.
Essentially such a tool would focus on the risk-factors for negative behaviour
associated with the experience of bullying. To identify what may be
considered general or global risk-factors a literature review was undertaken.
The review identified the following factors as commonly associated with
bullying:
• Excessive workload (Appelberg et al, 1961; Hoel & Cooper, 2000)
• Lack of control / lack of control of time (Zapf et al, 1996)
• Role conflict (Eianrsen et al, 1994; Vartia, 1996)
• Team-conflict (Hoel & Salin, 2003)
• Organisational politics (Kräkel, 1997; Salin, 2003)
• Organisational change (e.g. Neuman & Baron, 1998; Lewis & Sheehan,
2003)
• Perceived job insecurity/contingent workers (Knorz & Zapf, 1996;
Quinland, 1999)
• Organisational culture (Ashfort, 1994; Rayner et al., 2002)
However, the degree of empirical evidence in support of individual factors
varied, with for some factors contradicting emerging evidence. This
information was then related or merged with the result from the focus group
study outlined in the previous section. On this basis we developed a 26-item
risk-assessment tool which was then piloted as part of questionnaire
distributed to establish base-line measures across the participating
organisations (N=1033).
17
Although there was some evidence of the viability of our draft tool, the results
of the factors analysis of the pilot and a critical assessment of all items
making up the tool, suggested a need for further refinement. As a result a new
29-items instrument, the Bullying Risk Assessment Tool (BRAT) emerged. A
factor analysis (varimax rotation) yielded five factors. For a full overview of the
29 items instrument see Appendix B.
Extracts or quotations from the focus group discussions and examples of
items associated with each factors are provided below. (Please note that
some items making up the scales are worded positively and others negatively).
4.1 Factor structure
4.1.1 ‘Organisational fairness’
Sharing of work between colleagues was a common issue in a number of
organisations, particularly when in came to menial tasks:
“…My experience has been that you get the same people do them again
and again…they are the ones who do the extra bit and you’ll get other
people who will always find an excuse why they haven’t got the time to
do them.” NHS Ward Manager.
o I don’t know how to seek help within my organisation if I have a
problem with my line manager or supervisor
Grievance problems with managers were thought to be a particularly difficult
issue for some Civil Servants:
"Nobody knows where to go to if there’s something wrong. If you’ve
got a problem with your manager, nobody says “Right, you’ve got a
problem with your manager, you go to this person”. If you’ve got a
problem with your manager – tough! You know…there’s no guidelines,
there’s no checklist. If you say something like “Alright, I’ve got a
problem – I can’t relate to my manager”. Maybe I want to talk, your
manager’s a man or something – I want to talk to a female member of
staff. Where do I go?”
Other items:
o I feel my contribution to the organisation is recognised
o People in this organisation is not rewarded properly
4.1.2 ‘Team conflict’
o ‘Different professional groups don’t work well together within my
unit”
Conflict among team members was an issue raised in many groups.
Some participants emphasised that there was evidence of a lack of respect
between different professional groups. This was particularly apparent in the
18
NHS where nurses often experienced aggressive behaviour from doctors. The
quote below came from one nurse who was trying to make sense of a
Consultant’s behaviour by suggesting:
“Well, we get shouted at in theatre all the time by Consultants. And you
just realise that, it’s very impersonal. It’s nothing personal. It’s just that
they’re frustrated and it’s upsetting, but you don’t let it get in the way,
because the life of that person on the table, they are in charge of it, and
you just think “Right…he’s having a bad day. The life of the patient is
more important than your feelings”. We just get on with it.”
Other items:
o I don’t get on with some of my colleagues
o I find my colleagues to be cooperative
4.1.3 ‘Role conflict’
o I am not clear about how to carry out my job
Role conflict was another area frequently surfacing as suggested in the quote
below.
Other items:
o I face conflicting demands in my job
o My job description is clearly defined
4.1.4 ‘Workload’
o I feel that there isn’t enough hours in the day to complete my
work
Issues around workload were frequently raised by focus group participants:
A Clinical Manager exposed the constant pressure placed on them by the
demands of their job and in particular the struggle to schedule tasks within
a working day:
“I raised this issue with my line manager about that the amount and
expectation, that I will not be able to achieve the amount of work
that is passed down to my level, that I felt was unacceptable. I said
it’s impossible for me to do this, there were going to be things that
aren’t done and I have to prioritise what I can do and I can’t and the
answer he gave me back was that you’ll just have to do the best
you can. He said it’s what we all have to do, we have to do the best
we can and I said to him but I don’t think that’s good enough
because to me that is bullying, this is the Trust bullying me! I don’t
have 37 ½ hours written in my contract, it just says hours as
needed I forget the phrase but it says you know to do the job.”
19
Other items:
o Existing work pressure makes it difficult to take time off work
o Vacant positions are filled without any delay in my unit
4.1.5 ‘Leadership’
o I have confidence in my line-manager’s abilities
Leadership or lack of leadership was a frequently recurring theme:
“They don’t know how to react. All they know is that they come in, they
shut their door, they do their work, the jobs get allocated, the little mice
all run around and order everything and that’s it. But there’s no
interaction there. They don’t know how to interact with people because
nobody’s ever trained them to do that.” (Civil Service)
“I do think HR has become more removed from it, the emphasis is
moving much more, especially in this organisation which is relatively
new, moving towards line managers managing their staff and dealing
with situations but I’m not sure that a lot of people have got the
competencies and skills at the moment to do some of the things that
they’re asked to do or are asked to do and it’s all a little bit too much.
“ (Civil Service)
Other items:
o My line manager is sensitive to how I feel
o My line manager values constructive criticism
4.2 Factor reliability
Factor reliability was generally good (Chronbachs alpha of 0.57 – 0.81). The
lowest scale reliability emerged for the factor labelled ‘workload’. However, by
removing the item ‘my unit often makes use of temporary staff’, the alpha
increased to 0.62. It appears that this item might have a different meaning in
some of the organisations due to their particular practices regarding
temporary staff.
Table 1: BRAT factor reliability
Factor Scale reliability
Organisational fairness - OF 0.61
Team Conflict- TC 0.72
Role conflict – RC 0.65
Workload – WL 0.57
Leadership – LS 0.81
20
4.3 Validation of the BRAT
To validate the instruments a number of statistical tests were carried out. The
analysis revealed that the instrument can be considered valid in as much that
all factors, as well as the overall instrument, adequately emerged as
predictors of negative behaviour (measured by means of the revised Negative
Acts Questionnaire ((Einarsen & Hoel, 2001) and self-labelled bullying (in
response to a global definition of bullying) and with adverse consequences as
measured by means of mental health scores (General Health Questionnaire).
The NAQ-R is considered to measure two distinct factors, work-related
harassment and personal harassment (Einarsen & Hoel, 2001).
Table 2: Pearson correlations for BRAT scales
NAQ-R
Total
NAQ-R
WH
NAQ-R
PH
GHQ Bullying
BRAT – 29 item 0.581** 0.635** 0.450** 0.458** 0.419**
Org. fairness - OF 0.433** 0.463** 0.349** 0.315** 0.345**
Team Conflict- TC 0.467** 0.451** 0.386** 0.337** 0.416**
Role conflict – RC 0.406** 0.483** 0.278** 0.438** 0.203**
Workload – WL-6 0.296** 0.374** 0.183** 0.253** 0.168**
Workload – WL–5 0.304** 0.376** 0.194** 0.269** 0.166**
Leadership – LS 0.454** 0.478** 0.340** 0.315** 0.361**
** Correlation is significant at the 0.01 level
• All factors correlated significantly with total score on the NAQ-R, with
self labelled bullying and with total GHQ score
• Multiple regression indicated that all five risk factors are potentially
independent predictors of negative behaviour (R-sq=34.2%)
• Leadership followed by role conflict emerged as the single strongest
predictor of work-related harassment
• Leadership followed by team conflict emerged as the single strongest
predictor of personal harassment
• Multiple regression revealed that all factors predict work-related
harassment (R-sq=41.2%). Team conflict, leadership and
organisational fairness predict personal harassment (R-sq=22.3%)
• For self-labelled bullying – leadership, team-conflict, organisational
fairness and role conflict emerged as independent predictors (R-
sq=22.5%)
Results of the regression analysis are listed in Appendix C
It is worth noting that of those reporting themselves as having been bullied
within the last six months, 92% also reported having witnessed bullying talking
place within the same time period. This might have implications for application
of a risk-assessment tool as the far less sensitive issue of witnessing bullying
could to a large extent replace self-labelling whilst at the same time providing
similar information regarding predictors of bullying.
21
In order to assess the quality and potential effectiveness of the BRAT, the
scale’s test-retest reliability should be considered. However, one might
question whether it is feasible to successfully conclude on such an
instrument’s reliability.
Due to the sensitive nature of the issues under consideration, individuals are
likely to be influenced by daily events and there responses, therefore, could
be skewed. In other words, one might question to what extent risk-factors
could be expected to remain stable over time, especially where perceptions of
an individuals surroundings will at least to some extent fluctuate with task
requirement and ongoing dynamics of interpersonal interactions.
4.4 Application of the BRAT
It is important to consider applying the tool in its own right. In this respect it
was anticipated from the outset that the BRAT would be used in conjunction
with a more objective measure focussing on issues judged to be associated
with negative behaviour and bullying. The objectivity of this data refers to the
fact that it is taken from existing sources and is seen to be collated impartially
without any direct reference to bullying. For example, whether certain
changes have actually taken place or not, actual level of absenteeism
according to company records etc. For a further discussion of this, see section
7.
Although the risk-assessment tool can be used to identify individuals at risk of
bullying, the instrument is primarily aimed at establishing risk at a group-level.
In other words to what extent does some degree of consensus of risk exist at
a group level. This is important as one of the key outcomes of applying this
approach is to inform decision-making and highlight internal priorities. In this
respect organisational (or unit) action may be triggered by both overall scores,
relative scores on various factors or even on particular items. Returning to the
individual, where unit scores or risk is considered low whilst the scores of
particular individuals soar, one may conclude that there are other triggering or
contributing factors to bullying, possibly on a personal or one-to-one basis.
Such a conclusion would also require a different organisational response than
those cases where high risks are revealed at group level.
5. Identification and development of interventions
One of the aims of this research was: To develop and implement, according to
a particular research methodology, three different bullying intervention
programmes
This section provides details of the interventions and their rationale.
22
5.1 Policy communication
Whilst participating organisations are likely to have policies in place, the aims
of the policy communication intervention include raising awareness of the
organisation’s policy on bullying and the duty of organisational members in its
implementation. To ensure that the policies were in line with best practice in
the field, a set of guidelines were developed by the researchers on the basis
of close examination of a large number of such policies (see Appendix D).
Policies from each organisation were then carefully examined and compared
against it. Although differences were identified with respect to scope and
depth, all five organisations were seen to comply with our best practice
guidelines.
Rationale of intervention: It is envisaged that the individuals’ and others’
awareness of the presence of policies, the responsibility of managers with
regard to the implementation of policies and the potential consequences when
in breach of policy, would all impact on behaviour.
As the quote from a NHS Manager suggests the intention may not always be
to bully but the perception of the individual on the receiving end is important:
“It’s not about whether it happens or not sometimes, it’s about the
perception of the individual…when you listen to the staff, the victim,
and then they say this is how it made me feel, and you open your
policies and you think “gosh, yeah, it’s a perceptual thing!”
A lack of trust in grievance procedures was also highlighted on a number of
occasions:
“I just think, yeah, the organisation has a lot of responsibility and they
say a good job, they’ll have all the policies and they’ll say to you well
you can tick all these boxes and it is confidential but the absolute
reality is that it’s never confidential.”
The training includes:
• A statement of intent from senior managers highlighting the fact that
such behaviour will not be tolerated
• Outline of the managers / supervisors responsibility with regard to the
implementation of the policy and responsibility for challenging bullying
behaviour
• A definition of bullying and examples of bullying behaviour
• An overview of the complaints / grievance procedure and details of key
contact persons
5.2 Stress management training programme
This intervention follows the assumption that any negative and abusive
behaviour on the part of managers are frequently the result of work-overload
and the failure to deal with personal stress. The programme focuses on the
strain aspect of the stressor-strain relationship and the ability of individuals to
23
cope, and covers issues such as stress awareness and reduction. According
to some researchers, by becoming more aware of the relationship between
behaviour, personality, coping and stress outcomes, individuals may be better
able to cope with their stress. The aims of the stress management training
include raising awareness of stress and its impact on individuals and the
organisation, and developing manager / supervisor understanding of how to
manage their stress as well as the stress of people they are responsible for.
Rationale of intervention: by better controlling precursors of negative
behaviour resulting from stress, negative behaviour and bullying would be
reduced
Although bullying behaviour may not be accepted as part of an individuals’
character, a senior manager from the Police Force raises the possibility that
his behaviour could change adversely when under pressure:
“When I’m under pressure the aspiration is not to and I don’t know
whether I always control very well but I would think it would be a really
powerful person who under pressure kept so stable all the time. So
whether when I’m under pressure I change a little bit and maybe
become more prescriptive in what I want, which could be actually
construed as bullying, which I don’t think it probably would be because
I’m not of that ilk, but that is a possibility, but I try to stay calm but I
probably don’t always do it.”
The training includes:
• Defining stress
• Causes and consequences of stress
• Differing responses to stress
• How to identify stress in self and others
• Coping with stress (self) – problem focused coping, time management,
task prioritisation and exercise
• Managing stress in others
It is important to emphasise that the main aim of this intervention was to
assess whether stress management techniques have an effect on levels of
negative behaviour and presence of workplace bullying, rather than testing
the efficacy of stress management techniques in their own right.
5.3 Negative behaviour awareness training programme
Based on current knowledge and good practice guidelines in the area of
stress prevention and management which argue for researchers to target
interventions around organisational-specific issues, the purpose of this
intervention was to develop an evidence and needs-based programme to
tackle conflict and negative behaviour. Using feedback obtained from focus
groups and risk-assessment exercises, the aims of the negative behaviour
24
awareness training include raising awareness of negative behaviour and its
impact on individuals and the organisation, and developing a shared
understanding of what acceptable / unacceptable behaviour is within the
organisation.
Rationale of intervention: By raising awareness in a group context of negative
behaviour and bullying and providing participants with appropriate tools to
deal with difficult situations, bullying would be reduced.
Some extreme negative behaviours experienced and witnessed by a number
of people are carried out as a ‘right-of-passage’ or part of socialisation
processes and are seen as a bit of fun by perpetrators:
“Yes, well I had lots of practical jokes played on me in my earlier
years and some of them just weren’t funny, some of them were
indecent assaults by another name but they just thought it was quirt,
funny and all this sort of stuff…I had practical jokes played on me
which humiliated the life out of me and its that, it’s very personal, that’s
not a professional criticism, its about the person, I’m not just [NAME]
the Officer, it’s [NAME] the woman…
The training includes:
• Individual experiences of negative behaviour
• Definition of bullying and categories of bullying behaviour
• Evidence from previous research including effects on individual and
organisation
• Situations that cause bullying behaviour (organisation-specific
evidence from focus groups)
• Transactional analysis: how to develop skills for positive interaction
• A statement of intent from senior managers highlighting the fact that
such behaviour will not be tolerated
• Outline of the managers / supervisors responsibility to challenge
bullying behaviour
5.4 Implementation of interventions
The intervention programmes were all delivered by one person, who was an
independent and qualified trainer. A pilot of the interventions was carried out
in a non-participating NHS trust with the researchers present. On the basis of
our own observations and feedback from the participating managers (N=10)
each of the interventions were refined.
The interventions were evaluated by means of participant feedback obtained
at the end of the training session and six-months afterwards, comparison of
pre-intervention and post-intervention survey data and feedback from post-
intervention focus groups.
25
6. Questionnaire survey participants and sample background
6.1 Sample - study participants
We requested assistance from all five participating organisations in the
distribution of approximately 100 questionnaires to a random sample of
employees from each of the five intervention groups (approximately 500
questionnaires in total in each organisation) on two occasions – pre-
intervention for baseline measure purposes and then approximately 6 months
post-intervention for evaluation purposes. Questionnaires were sent out by
participating organisations accompanied by a cover letter from them outlining
the study, introducing the research team, and assuring confidentiality and
anonymity. A stamped addressed envelope was provided for questionnaires
to be returned directly to researchers at the University of Manchester.
In general the questionnaires were completed in an acceptable manner, with
little data missing for any variable. Of the 2505 questionnaires sent out during
the pre-intervention survey 1041 were returned to the researchers,
representing an overall response rate of 41.5 %. A total of 884 questionnaires
from the 2499 sent out during the post-intervention survey were returned to
the researchers, representing an overall response rate of 35.4%. Details of
both the pre-intervention and post-intervention questionnaire surveys are
presented in table (3) below.
Table 3: Survey sample
Organisation Total
sample
No. of
questionnaires
Returned
Response
rate (%)
T(0) 508 260 51.2 Civil Service Dept.
T(1) 499 249 50.0
T(0) 497 185 37.2 NHS Mental Health
Trust (NHS 1)
T(1) 500 162 32.4
T(0) 500 142 28.4 Acute NHS Trust –
North (NHS 2)
T(1) 500 160 32.0
T(0) 500 212 42.4 Acute NHS Trust –
South (NHS 3)
T(1) 500 139 27.8
T(0) 500 242 48.4 Police Force
T(1) 500 174 34.8
T(0) 2505 1041 41.6 Total Sample
T(1) 2499 884 35.4
26
In addition to distributing the questionnaire survey to employees,
organisations were asked to respond to the objective part of our risk
assessment tool by submitting information on each of the five groups
participating in this research.
Demographic details of the complete sample and their experiences of bullying
are presented below.
6.2 Sample background
The respondents were asked to ‘tick the box that best describes yourself or
your situation’.
6.2.1 Gender
Although there were substantially more female respondents, particularly within
NHS organisations, the proportion of female and male employees during both
surveys remained approximately the same. The gender breakdown within our
sample reflects the nature of organisations involved.
Table 4: Sample gender breakdown
Gender
(%)
Pre
Post
Int
Civil
Service
NHS 1 NHS 2 NHS 3 Police Total
Sample
T(0) 46.5 22.8 24.8 19.0 57.2 36.2 Male
T(1) 52.0 28.8 20.5 13.1 53.8 36.4
T(0) 53.5 77.2 75.2 81.0 42.8 63.8 Female
T(1) 48.0 71.2 79.5 86.9 46.2 63.6
Pre-intervention
Gender
Female
63.8%
Male
36.2%
Post-intervention
Gender
Female
63.6%
Male
36.4%
27
6.2.2 Age
The mean age of respondents was 43 years at both survey points. The
majority of participants in both the pre and post-intervention surveys came
from the 35-44 and 45-54 age groups on both occasions, although employees
from the Civil Service organisation were evidently younger. The sample did
not differ significantly between the two periods.
Table 5: Sample age breakdown
Age
(%)
Pre
Post
Int
Civil
Service
NHS 1 NHS 2 NHS 3 Police Total
Sample
T(0) 40.61 45.18 43.68 43.13 42.88 42.84 Mean
T(1) 41.07 46.45 43.02 44.69 43.17 43.39
T(0) 6.0 4.1 3.6 2.5 4.0 4.2 16-24
T(1) 6.9 2.0 4.1 3.1 2.4 4.0
T(0) 28.2 12.3 15.9 17.5 17.9 19.2 25-34
T(1) 27.9 9.3 14.3 15.3 18.2 18.1
T(0) 27.4 25.7 34.8 31.5 31.3 29.9 35-44
T(1) 26.2 26.5 36.7 24.4 30.9 28.8
T(0) 25.0 39.2 28.3 35.0 32.6 31.7 45-54
T(1) 22.7 43.7 29.3 41.2 33.9 32.9
T(0) 13.3 18.7 17.4 13.5 14.3 15.1 55-70
T(1) 16.3 18.5 15.6 16.0 14.5 16.2
0%
5%
10%
15%
20%
25%
30%
35%
Pre-intervention 4.2% 19.2% 29.9% 31.7% 15.1%
Post-intervention 4.0% 18.1% 28.8% 32.9% 16.2%
16-24 yrs 25-34 yrs 35-44 yrs 45-54 yrs 55-70 yrs
28
6.2.3 Ethnicity
Although percentages for specific ethnic categories were low at both measure
points, an above national average figure of more than one in ten respondents
were from a Black and Minority Ethnic (BME) background (assuming that the
‘other’ ethnicity group classify themselves as ‘non-white’).
Table 6: Sample ethnicity breakdown
Ethnicity Percent:
pre-intervention
Percent:
post-intervention
White 88.7 89.3
Non-white: 11.3 10.7
Mixed 1.0 0.7
Indian 1.5 1.4
Pakistani 0.4 0.2
Bangladeshi 0.3 0.2
Chinese 0.6 0.7
Other Asian 1.7 1.6
Caribbean 2.1 1.2
African 2.1 3.0
Other Black 0.4 0.2
Other 1.4 1.4
6.2.4 Contracted hours of work
The majority of respondents in both our surveys reported to working full-time.
Approximately one in five employees reported to working part-time hours
during the pre and post-intervention surveys.
79.7
81.8
20.3
18.2
0
10
20
30
40
50
60
70
80
90
Full-time % Part-time %
Pre-intervention -
Contracted hours
Post-intervention -
Contracted hours
29
6.2.5 Employment contract
As suggested in bar chart below, the vast majority of participants were
employed on permanent contracts. The numbers for the other categories were
relatively small.
95.3
93.6
2.5 2.8
1.2 1.5 1.1
2.1
0
10
20
30
40
50
60
70
80
90
100
Permanent % Fixed period or
task %
Seconded % Other %
Pre-intervention - Contracted hours
Post-intervention - Contracted hours
6.2.6 Job type
Almost half the sample came from ‘professional / managerial’ grades. The
sample did not differ significantly between the pre and post intervention
survey period.
Table 7: Sample job type breakdown
Job type Percent:
pre-intervention
Percent:
post-intervention
Clerical / admin 22.7 23.4
Professional / managerial 49.7 49.1
Specialist / technical 16.8 18.0
Other 10.9 9.4
6.2.7 Supervisory responsibilities
Almost half of the respondents suggested that their job did not involve any
managerial responsibilities whatsoever. Approximately the same number
reported having a supervisory or managerial role. In this respect the sample
remained relatively constant between the pre and post intervention survey
period.
30
Table 8: Sample supervisory responsibilities breakdown
Supervisory responsibilities Percent:
pre-intervention
Percent:
post-intervention
No supervisory responsibilities 46.7 45.9
Supervisor 22.0 21.1
Middle manager 18.6 19.0
Senior manager 6.1 6.8
Other 6.6 7.2
6.2.8 Length of time with organisation
A large number of respondents were employed within their current
organisations for more than ten years, reflecting a relatively stable
employment relationship. The figures remained almost constant over the
intervention period.
8.1
6.6
21.2 21.3
15.9
17.3
10.7
9.5
44.2
45.2
0
5
10
15
20
25
30
35
40
45
50
Under 1 year 1-3 yrs 4-6 yrs 7-9 yrs 10 or more yrs
Pre-intervention - Time with
organisation
Post-intervention - Time with
organisation
7. Objective data
In line with the intervention literature’s suggestions regarding methodological
progress, it was an expressed intention of the study protocol to develop a
procedure by which management could systematically measure certain
organisational indicators believed to be influenced by the presence of risk
factors of negative behaviour and bullying. Compared to self-reported data
obtained by means of questionnaires, such data can be considered more
objective in the sense that they are not open to cognitive processing to the
same degree but simply report on recorded organisational data.
With reference to bullying literature we developed a report form consisting of a
series of questions on measures such as absenteeism, turnover,
grievances/complaints and various forms of changes potentially taking place
31
within the organisation, e.g. restructuring, financial cutback and change of
managers. To provide an accurate account of these measures without any
potential interference from line-management, these forms were to be filled in
by the human resources department. An example of the form is provided in
Appendix E.
Unfortunately, despite informing the participating organisations about the
need to collect such data and the format by which this would be undertaken,
several organisations had great difficulties in obtaining the data in the form
required. In some cases data were only available at an organisational level
and not at the unit level needed to measure any potential impact of the
interventions. Thus, incomplete records on several measures particularly at
T1 (six months after the intervention) largely undermined the opportunity to
validate self-report measure. Similarly, incomplete records would also
undermine the opportunity to successfully apply these measurements to the
evaluation process. We report on the findings where we have complete data
from both the pre-intervention and post-intervention in section 10.
As we needed a uniform approach to the collection of ‘objective measures’ in
order to be able to compare data across organisations, we saw it as our
responsibility to identify the measures and develop the forms and procedures
needed to collate them in a comparable format. However, with the benefit of
hindsight it may be argued that management (and steering committees) in
participating organisations could have been more involved in the process, in
particular with the identification of valuable organisational measures as well
as reviewing internal data collection processes at an earlier stage to ensure
that they could comply with our requests. In other words, potential problems
could have been identified and rectified at an early stage by widening their
input to the process and by testing the procedure.
8. Experiences of bullying in the workplace: baseline measures
This section reports on the main baseline measures taken prior to
implementation of the intervention programme. For comparison purposes, we
also present findings from our previous BOHRF study (Hoel & Cooper, 2000).
The section reports only on the overall findings. For a further breakdown of
findings for different demographic groups, see Appendix F. This is particular
relevant since several measures applied in the current study were the same
as those used for the nationwide study reported in 2000.
8.1 Experience of bullying
Before answering questions associated with experience of bullying,
respondents were presented with the following statement/definition:
We define bullying as a situation where one or several individuals
persistently over a period of time perceive themselves to be on the
receiving end of negative actions from one or several persons, in a
situation where the target of bullying has difficulty in defending him or
32
herself against these actions. We will not refer to a one-off incident as
bullying.
‘Using the above definition, please state whether you have been bullied at
work over the last 6 months’.
A total of 13.6% of respondents report that they have been bullied within the
last six months. The persistency of their experience is illustrated by the table
below.
Table 9: Percentage of respondents reporting to being bullied
No Yes
(Total)
Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-
intervention
86.4 13.6 3.2 5.9 2.6 1.4 0.6
BOHRF 1 89.4 10.6 1.9 6.2 1.0 0.8 0.6
Pre-intervention
No
86.4%
Yes
13.6%
These figures suggest that a substantial proportion of employees in the
participating organisations considered themselves to being bullied, thus
entirely justifying the attention to the issue and current intervention study.
Moreover, compared to the nationwide study the overall exposure rate is
higher. Looking at exposure rates for individual organisations, it is interesting
to compare them with occupational/industry average figures in the BOHRF (I)
- study which were 9.9%, 10.6% and 12.1% for the Civil Service, the NHS and
the Police respectively. As we don’t know if our samples are representative of
their respective organisations or indeed of their occupation/industrial sector,
we cannot say that there has been a rise in exposure rates over the last five
year. However, what we can say is that the problem certainly appears to
affect a substantial part of the workforce and remains an important challenge
across organisational settings.
33
Table 10: Percentage of respondents from each organisation reporting to
being bullied - baseline
8.2 Witnesses of bullying
A total of 44% reported that they had observed or witnessed bullying taking
place within the last 6 months, indicating that nearly half of all respondents
would have either directly or indirectly experienced the problem.
Table 11: Percentage of respondents reporting to witnessing bullying in the
last 6 months
8.3 Past bullying
More than a quarter of the sample who said they weren’t being bullied at
present had been bullied in the last 5 years. This is mainly in line with findings
from the previous study.
Table 12: Percentage of respondents reporting to experiencing bullying in the
past 5 years
No Total
Yes
Yes,
very
rarely
Yes, now
and then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Civil Service 87.5 12.5 3.9 4.3 1.6 1.6 1.2
NHS 1 89.2 10.8 2.8 3.4 0.6 2.8 1.1
NHS 2 84.1 15.9 2.2 8.0 5.1 0.7 0
NHS 3 83.4 16.6 3.4 7.8 3.9 1.5 0
Police 86.9 13.1 3.0 6.8 2.5 0.4 0.4
No, Never
(%)
Total
Yes (%)
Yes, but
rarely (%)
Yes, now
and then %)
Yes, often
(%)
Pre-intervention 56.0 44.0 25.8 13.9 4.3
No (%) Yes (%)
Pre-intervention 72.3 27.7
BOHRF 1 75.3 24.7
34
8.4 Targets of bullying
8.4.1 Gender
A slightly greater proportion of women (13.9%) reported being bullied
compared to men (12.3%). However, this difference was not significant.
Table 13: Gender differences in exposure to bullying
Female (%) Male (%) Sig.
Pre-intervention 13.9 12.3 NS
BOHRF 1 11.4 9.9 NS
0%
2%
4%
6%
8%
10%
12%
14%
16%
Pre-intervention 13.9% 12.3%
BOHRF 1 11.4% 9.9%
Female Male
8.4.2 Age
With the exception of the youngest respondent group, only minor differences
emerged when comparing bullying for different age groups. However,
compared to the nationwide study the younger respondents reported less
bullying whilst those over fifty years of age appear to be more frequently
exposed.
Table 14: Age differences in exposure to bullying
16-24
(%)
25-34
(%)
35-44
(%)
45-54
(%)
55-70
(%)
Sig.
Pre-intervention 4.9 13.4 14.0 12.3 15.9 NS
BOHRF 1 14.0 10.5 13.0 12.0 8.8 NS
35
0
2
4
6
8
10
12
14
16
18
Pre-intervention 4.9 13.4 14 12.3 15.9
BOHRF 1 14 10.5 13 12 8.8
16-24 (%) 25-34 (%) 35-44 (%) 45-54 (%) 55-70 (%)
8.4.3 Full-time versus part-time
As in the nationwide study, respondents in full-time jobs were more likely to
be bullied than someone in a part-time job at a level bordering significance,
with 13.9% of full-timers as opposed to 8.9% of part-timers being bullied.
Table 15: Working hours differences in exposure to bullying
8.4.4 Ethnicity
Respondents of ethnic groups other than white were substantially more likely
to be bullied than white respondents, with 25.2% of non-white respondents as
opposed to 11.8% whites (p<0.001).
0%
5%
10%
15%
20%
25%
30%
Pre-intervention
BOHRF 1
Pre-intervention 11.8% 25.2%
BOHRF 1 10.5% 14.1%
White Non-white
Full-time (%) Part-time (%) Sig.
Pre-intervention 13.9 8.9 p=0.066
BOHRF 1 14.5 7.1 P<0.001
36
8.5 Organisational level of target
The current study confirms our previous findings (Hoel & Cooper, 2000) that
bullying is a problem across organisational hierarchies with just minor
differences emerging between the organisational levels or status groups.
Table 16: Organisational level of target
No
supervisory
responsibility
(%)
Spvsr.
(%)
Middle
Mgmnt.
(%)
Snr.
Mgmnt.
(%)
Other
(%)
Pre-intervention 13.5 13.0 12.6 11.9 18.0
BOHRF 1 9.6 9.1 10.6 8.5 11.0
8.6 Perpetrators of bullying
As evident from table 17 below, over 60% of bullying victims reported that
they were bullied by someone in a supervisory or managerial capacity. These
figures could suggest that there is a slight decrease in ‘vertical bullying’ since
our nationwide study. However, closer inspection of the nationwide study
showed that the NHS also had a lower percentage of vertical (top-down)
bullying (64.3%).
Table 17: Perpetrators of bullying
Table 18: Perpetrators of bullying: breakdown within organisations
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Pre-intervention 61.3 42.3 9.5 14.6
BOHRF 1 74.7 36.7 6.7 7.8
Spvsr /
manager(s)(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Civil Service 71.9 37.5 9.4 6.3
NHS 1 47.4 52.6 10.5 10.5
NHS 2 57.1 47.6 4.8 14.3
NHS 3 47.1 47.1 14.7 32.4
Police 77.4 32.3 6.5 6.5
37
8.7 How many were bullied?
Again the findings were very similar to the nationwide survey with just under a
third a third of respondents (29.9%) report to have been bullied on their own
or singled out for bullying, although substantial discrepancies exist between
the organisations. A total of 17.2% reported that everyone in their work group
were bullied, representing a minor increase compared to the BOHRF I study.
A breakdown of these findings for each organisation is provided in the table
19 below.
Table 19: How many were bullied
Table 20: How many were bullied: breakdown within organisations
8.8 Duration of the bullying experience
Approximately 60% of those who reported being bullied had been exposed to
bullying for more than a year, and approximately a third for more than 2 years.
Table 21: Duration of bullying experience
8.9 Experiences of negative behaviours
‘Someone withholding information which affects your performance’ followed
by ‘having your views and opinions ignored’ were the two most frequently
experienced negative behaviours. Amazingly, looking at the order by which
these behaviours were experienced it emerged that the ranking of the
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Pre-intervention 29.9 53.0 17.2
BOHRF 1 31.2 54.9 14.8
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Civil Service 31.3 46.9 21.9
NHS 1 38.9 55.6 5.6
NHS 2 15.0 55.0 30.0
NHS 3 32.4 58.8 8.8
Police 30.0 50.0 20.0
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Pre-intervention 20.0 21.5 25.9 32.6
BOHRF 1 16.8 16.4 27.5 39.3
38
behaviour was a near replica of the order emerging from the nationwide study
with only two behaviours changing place. This strongly suggests that although
organisational differences exist, both studies provide an accurate picture of
the nature of the behaviours experienced.
Table 22: Experiences of negative behaviour
*(R)egularl y or (O)ccasionally experiences
8.10 Outcomes of bullying
In line with previous research the correlations in table 23 below suggest that
bullying is relatively strongly correlated with mental health. Among other
outcomes often associated with bullying and with direct implications for the
organisation, the correlation with intension to quit is particularly strong. In both
cases the strength of correlations is substantially above those found in the
nationwide survey which were 0.267 and 0.212 respectively.
Table 23: Negative Acts: Breakdown within organisations
*
Pre-intervention (%) BOHRF 1 (%)
O 47.0 54.0 Someone withholding information which affects
your performance
R 10.0 13.3
O 40.4 49.3 Having your opinions and views ignored
R 8.2 7.8
O 33.8 39.3 Being exposed to an unmanageable workload
R 13.7 14.6
O 31.0 35.1 Being ordered to do work below your
competence
R 9.4 10.7
O 29.2 42.2 Being given tasks with unreasonable or
impossible targets or deadline
R 8.7 9.7
O 24.5 32.0 Having key areas of responsibility removed or
replaced with more trivial or unpleasant tasks
R 6.9 6.1
O 22.5 29.8 Spreading gossip
R 4.8 4.1
O 22.4 25.1 Being shouted at or being the target of
spontaneous anger R 4.7 4.7
O 22.3 27.8 Being humiliated or ridiculed in connection with
you work
R 4.6 3.6
Outcome measures
Have you been bullied at work
over the last 6 months?
p
Mental Health (GHQ 12) 0.392 <0.001
Intention to quit 0.328 <0.001
Job satisfaction -.178 <0.001
Psychological contract -.257 <0.001
39
9. Experiences of bullying in the workplace pre and post intervention
comparisons
By comparing baseline and post-intervention data, this section reports on the
main changes associated with bullying and negative behaviour which have
occurred within the duration of this study. For a breakdown of further pre- and
post-intervention, see Appendix G.
9.1 Experiences of bullying
During the baseline measure, a total of 13.6% of respondents reported to
being bullied in the previous six months. This measure increased slightly to
14.3% post-intervention (see table 24 below).
Table 24: Experiences of bullying – pre and post intervention
Pre-intervention
No
86.4%
Yes
13.6%
Post-intervention
No
85.7%
Yes
14.3%
As highlighted in table 25 below, experiences of bullying remained relatively
stable within all the organisations between the pre and post-intervention
period, except for a marked increase in experiences within NHS (3) from
16.6% to 23%, and a decrease in experiences in the Police Force from 13.1%
to 11.8%.
No Total
Yes
Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-intervention 86.4 13.6 3.2 5.9 2.6 1.4 0.6
Post-intervention 85.7 14.3 3.7 7.4 1.2 1.0 0.9
40
Table 25: Organisational breakdown of experiences of bullying – pre and post
intervention
9.2 Perpetrators
With a small reduction in numbers for supervisors / managers as perpetrators
of bullying, it is tempting to suggest that our interventions, which were
predominantly targeted at this group, could have had an effect on their
behaviour. However, the increase in numbers of colleagues who are bullying
(table 26 below) is more difficult to explain as is the small reduction in bullying
from clients / customers / patients.
Table 26: Perpetrators of bullying
As suggested in table 27 below, the reduction in bullying from managers /
supervisors was experienced in three of the five organisations involved, with a
marked decrease in NHS (2) (down to 45.8% from 57.1%) and the Police
Force (down to 65% from 77.4%). All organisations, except NHS (2) reported
a substantial reduction in experiences of bullying from clients.
No Total
Yes
Yes, very
rarely
Yes,
now and
then
Yes, several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Civil Service (T0) 87.5 12.5 3.9 4.3 1.6 1.6 1.2
Civil Service (T1) 87.4 12.6 3.6 4.5 1.2 1.2 2.0
NHS 1 (T0) 89.2 10.8 2.8 3.4 0.6 2.8 1.1
NHS 1 (T1) 89.2 10.8 4.5 3.2 1.9 0.6 0.6
NHS 2 (T0) 84.1 15.9 2.2 8.0 5.1 0.7 0
NHS 2 (T1) 84.3 15.7 2.0 11.8 0 2.0 0
NHS 3 (T0) 83.4 16.6 3.4 7.8 3.9 1.5 0
NHS 3 (T1) 77.0 23.0 7.4 13.3 0.7 0 1.5
Police (T0) 86.9 13.1 3.0 6.8 2.5 0.4 0.4
Police (T1) 88.2 11.8 1.8 7.1 1.8 1.2 0
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Pre-intervention 61.3 42.3 9.5 14.6
Post-intervention 57.4 47.5 9.8 9.0
41
Table 27 Organisational breakdown of perpetrators of bullying
9.3 How many were bullied?
As highlighted in table 28, the number of people who are singularly bullied has
increased from 29.9% pre-intervention to 36.1% post-intervention. At the
same time there is a substantial decrease in work group bullying from 17.2%
to 10.9%.
Table 28: How many were bullied
A breakdown of the data by organisations (table 29 below) also highlights
organisational variations in the number of people bullied, but with an overall
trend of a reduction in bullying of ‘everyone in the work group’.
Spvsr /
manager(s)
(%)
Colleague(s)
(%)
Subordinate(s)
(%)
Client(s)
(%)
Civil Service (T0) 71.9 37.5 9.4 6.3
Civil Service (T1) 66.7 33.3 10 3.3
NHS 1 (T0) 47.4 52.6 10.5 10.5
NHS 1 (T1) 52.9 64.7 5.9 5.9
NHS 2 (T0) 57.1 47.6 4.8 14.3
NHS 2 (T1) 45.8 50.0 25.0 16.7
NHS 3 (T0) 47.1 47.1 14.7 32.4
NHS 3 (T1) 54.8 51.6 3.2 12.9
Police (T0) 77.4 32.3 6.5 6.5
Police (T1) 65.0 45.0 5.0 5.0
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Pre-intervention 29.9 53.0 17.2
Post-intervention 36.1 52.9 10.9
42
Table 29: Organisational breakdown of how many were bullied
9.4 Psychological contract
The term ‘psychological contract’ is being increasingly used in organisational
studies. It suggests that reciprocal exchanges enable individuals to control the
giving and taking that is involved in work settings without the development of
feelings of injustice. Employees observe their working relationships as a form
of social exchange by contributing effort and loyalty in return for discernible
benefits such as pay and recognition (Eisenberger, Huntingdon, Hutchinson &
Sowa, 1986). The ‘psychological contract’ is an implicit agreement between
the employer and employee that they will treat each other fairly. Although it is
not legally binding, both parties can aspire to maintain mutual trust and
potentially build a strong relationship. However, as it constitutes an emotional
bond, if it is broken it can lead to the feeling of betrayal and resentment
(Rousseau, 1995).
We used Sandra Robinson’s (1996) 7-item measure of psychological contract
fulfilment to evaluate the nature of employee psychological contracts. As
indicated in the table below (table 30) the figures suggest an overall negative
trend in psychological contract fulfilment in relation to experiences of bullying.
Table 30: The impact of experiences of bullying on victims’ psychological
contracts
Only you
(%)
You and several
other colleagues (%)
Everyone in your
work group (%)
Civil Service (T0) 31.3 46.9 21.9
Civil Service (T1) 44.8 41.4 13.8
NHS 1 (T0) 38.9 55.6 5.6
NHS 1 (T1) 52.9 41.2 5.9
NHS 2 (T0) 15.0 55.0 30.0
NHS 2 (T1) 26.1 65.2 8.7
NHS 3 (T0) 32.4 58.8 8.8
NHS 3 (T1) 25.8 64.5 9.7
Police (T0) 30.0 50.0 20.0
Police (T1) 36.8 47.4 15.8
Mean
psychologic
al contract
score
Sample
mean
No Yes Yes,
very
rarely
Yes,
now
and
then
Yes,
several
times a
month
Yes,
several
times a
week
Yes,
almost
daily
Pre-
intervention
22.03 22.59 18.46 20.86 18.18 16.92 18.69 15.40
Post-
intervention
21.66 22.35 17.77 19.66 18.44 13.70 14.13 14.38
43
10. Efficacy of intervention: comparing baseline and post-intervention
measures for different interventions across organisations
In order to establish whether any changes have taken place in the period
between when base-line measures were obtained and six-months following
the interventions, a univariate analysis of variance was carried out on a
number of variables that could potentially be affected by the interventions.
These results are based upon an analysis of two sets of questionnaire
responses: Out of the 2,505 questionnaires distributed prior to the
interventions (baseline) a total of 1,041 were returned, representing a
response rate of 41.5%. 2,499 questionnaires were distributed for the post-
intervention (t1) measurement from which 884 were returned, representing a
reduced response rate of 35.4%. Overall these figures can be considered very
satisfactory for these types of studies.
The intervention groups referred to below are numbered as follows:
1 Control group
2 Policy communication (only)
3 Stress (and policy communication)
4 Negative behaviour awareness (and policy communication)
5 All (Stress, Negative behaviour awareness & policy communication)
10.1 Overall results
A series of analyses revealed that whilst a significant difference was found
between baseline scores and post-intervention scores for some variables
within all five groups, there was no significant difference for any of the key
variables between the experiment groups with respect to any change between
baseline and post-intervention scores. Overall scores for each intervention
group (scores across the five organisations) have been included in the tables
below and are introduced at the bottom of each table.
To examine the potential impact of different interventions in each of the
participating organisations, we carried out a univariate analysis of variance
(Test of between subjects effect). Overall no statistically different results
emerged for any of the key variables. Whilst in some cases there were
substantial differences between interventions, these differences were not
systematically upheld across all five organisations. It is, therefore, impossible
at this stage to conclude that any particular intervention or combination of
interventions is more effective than any other. However, in order to elucidate
trends in the data a closer examination of key variables was undertaken. The
data for each of these variables are presented and discussed below.
44
10.2 Bullying
Table 31: Bullied within the last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 baseline 1.21 1.27 1.20 1.41 1.42
T1 1.15 1.31 1.28 1.31 1.50
2 baseline 1.09 1.32 1.18 1.18 1.83
T1 1.24 1.17 1.38 1.06 1.23
3 baseline 1.50 1.17 1.25 1.24 1.65
T1 1.38 1.25 1.31 1.06 1.65
4 baseline 1.26 1.49 1.40 1.36 1.30
T1 1.35 1.31 1.29 1.69 1.45
5 baseline 1.29 1.36 1.15 1.34 1.22
T1 1.16 1.50 1.09 1.38 1.22
All orgs. Baseline 1.25 1.33 1.23 1.32 1.46
T1 1.23 1.31 1.27 1.31 1.43
Higher scores suggest increased levels of bullying
The above table indicates that very minor changes have taken place during
the six-month period between baseline and post-intervention measurements.
With more than 85% of respondents answering no (1) to this question, as well
as given the way the question is phrased which covers experience around the
time of the interventions this may be expected.
To be able to spot potential trends in the relative change between the base-
line and post-intervention (T1) measures, we simplified the above table by
means of showing increases in scores as +; decreases as -; and no changes
as 0. The use of plus (+) and minus (-) only refers to whether there is a
numerical increase (+) or decrease (-) in the scores for the instrument applied
and does not indicate whether the change is in the desired direction or not.
However, to avoid confusing the reader the desired direction is highlighted for
each table.
The number of pluses or minuses reflects the degree of change in a particular
direction, with each plus (+) or minus (-) representing a 5% change from the
baseline measure. In calculating the relative change between baseline and
post-intervention measures the minimum score on any scale is taken into
consideration. This re-scaling is done in order to avoid underestimating any
change.
However, due to the way bullying was measured with the large majority of
respondents answering ‘never’, the table is somewhat skewed. The
calculation of relative changes is, therefore, based on absolute values as
showed in the table. Each plus (+) and minus (-) represents an absolute
change of 0.1.
45
Table 32: Trends in experience of bullying
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 0 - 0
2
+ - ++ - - - - - - -
3
- 0 0 - 0
4
0 - - +++ +
5
- + 0 0 0
Higher scores (+) suggest an increase in levels of bullying
10.3 Witnessing bullying
It may be useful to look at equivalent scores for those who have witnessed
bullying as they would account for a much larger proportion of the
respondents than targets themselves and thus, possibly give a more accurate
picture of the situation.
Table 33: Witnessing bullying over the last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 baseline 1.49 1.67 1.48 1.51 1.47
T1 1.35 1.80 1.41 1.62 1.78
2 baseline 1.56 1.48 1.65 1.68 2.03
T1 1.55 1.47 1.51 1.50 1.42
3 baseline 2.00 1.66 1.63 1.80 1.90
T1 2.07 1.73 1.46 1.44 1.82
4 baseline 1.61 1.82 1.89 2.02 2.00
T1 1.92 1.69 1.81 1.78 1.92
5 baseline 1.67 1.60 1.54 1.43 1.43
T1 1.55 1.64 1.29 1.59 1.39
All orgs. Baseline 1.63 1.66 1.62 1.69 1.73
T1 1.63 1.69 1.49 1.59 1.67
Higher scores suggest an increase in witnessing bullying
To be able to picture any changes between base-line and post-intervention,
the same exercise as carried out for the previous variable also apply to the
variable ‘witnessing bullying’.
46
Table 34: Trends in witnessing bullying over the last six months
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
- + 0 + +++
2
0 0 - - - - - - - -
3
0 0 - - - - 0
4
+++ - 0 - - 0
5
- 0 - - + 0
Higher scores (+) suggest an increase in witnessing bullying
The above figures reinforce the view that no clear conclusion can be drawn
with regard to effectiveness of a particular intervention. However, with the
exception of organisation 1, there appears to be a slight downwards trend for
all groups which have received some training. Given that training and other
attention to workplace problems would tend to sensitise participants to the
problem, this overall slight reduction could signal a small reduction in levels of
bullying.
10.4 Negative behaviours
Table 35: Negative Acts Questionnaire (NAQ-R) - Minimum score on scale =
22
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 26.78 28.62 26.96 27.31 28.97
T1 26.82 27.83 28.55 27.10 29.94
2 Baseline 27.70 32.29 30.00 26.90 30.59
T1 27.07 30.69 33.28 28.47 28.08
3 Baseline 31.44 28.31 27.75 27.14 32.97
T1 31.88 28.81 28.30 26.79 28.29
4 Baseline 31.00 28.29 28.68 29.15 30.61
T1 31.22 28.48 27.38 29.36 32.10
5 Baseline 28.13 29.78 27.70 27.81 28.76
T1 28.17 30.03 27.19 27.81 29.41
All orgs. Baseline 28.57 29.29 28.13 27.76 30.25
T1 28.63 28.96 29.23 27.89 29.30
Higher scores suggest an increase in reported negative behaviour
Again, to be able to spot potential trends in relative change between baseline
and post-intervention (T1) measures we applied the same procedure as for
self-labelled bullying (see above).
47
Table 36: Trends in Negative Acts (NAQ-R) Minimum score on scale = 22
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 - - ++++++ 0 ++
2
- - - - - ++++++
+++
++++
++
- - - - - -
3
+ + ++ - - - - - - -
- -
4
0 0 - - - - 0 ++++
++
5
0 0 - 0 ++
Higher scores suggest an increase in reported negative behaviour
The above table indicates that in the case of negative behaviour also, no
systematic changes appear to have occurred between baseline measurement
and the post-intervention measurement. With only approximately 10-20% of
employees in the units trained (primarily supervisors / managers), it is of
course possible that the amount of training or number of people trained is
insufficient to affect behavioural patterns. Although the effect of training
generally is expected to be greatest straight after the training has taken place,
it is also of course possible that a six months delay between interventions and
measurements might have been too short for any changes to occur and be
noted. Alternatively, different interventions may in some cases have different
effects in different organisational settings, with other external or internal
circumstances playing a part. For example, organisation 1 experienced
unforeseen changes in the time between baseline and post-intervention
measures were obtained, with the possibility of redundancy affecting a
substantial number of employees taking part in the study.
10.5 The psychological contract
When we compare the psychological contract of individuals from the five
interventions groups, our data suggests that there is an overall reduction in
psychological contract fulfilment scores except for individuals who took part in
the full day training session (group 5). It was, therefore essential to check
whether these findings were generally upheld when looking at the impact of
individual interventions. The same procedure used for bullying and negative
behaviour was again applied.
48
21.69
21.96
22.02
23.12
21
20.88
21.19
21.96
22.9
21.19
20.5
21
21.5
22
22.5
23
23.5
Group 1 Group 2 Group 3 Group 4 Group 5
Pre-intervention Post-intervention
Table 37: Psychological contract – (Minimum score on scale = 7)
Intervention group mean scores
Organisation
Time of
assessment
1 2 3 4 5
1 Baseline 23.21 21.77 22.04 21.43 20.71
T1 22.57 21.38 21.34 22.27 19.64
2 Baseline 21.78 21.67 21.41 24.21 22.87
T1 20.79 21.78 20.62 24.25 22.17
3 Baseline 18.79 21.50 21.65 22.37 20.41
T1 17.95 20.13 21.14 23.89 22.44
4 Baseline 20.44 23.23 23.07 23.91 17.70
T1 20.05 21.35 24.21 21.94 20.50
5 Baseline 22.40 21.51 21.64 24.13 22.18
T1 21.48 21.23 22.00 22.52 21.74
All orgs. Baseline 21.70 21.96 22.02 23.15 20.86
T1 20.95 21.16 21.78 22.86 21.17
Higher scores suggest an improvement in the psychological contract
Table 38: Trends in Psychological contract (Minimum score on scale = 7)
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 0 + -
2
- 0 - - 0 0
3
- - 0 ++ +++
4
0 - - + - - +++++
5
- 0 0 - 0
Higher scores suggest an improvement in the psychological contract
49
Again it was impossible to trace any clear patterns in the data. Since
increased scores in this case represent an improvement of the psychological
contact, with a few exceptions improvement of the psychological contract was
associated with training in one form or another, with the most positive
development associated with the full day training (all training).
10.6 Mental health
Table 39: Mental health measure by GHQ-12 - Minimum score on scale = 12
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 24.15 25.07 24.29 24.29 24.26
T1 24.77 24.79 25.00 23.86 24.88
2 Baseline 24.37 26.42 25.69 23.87 25.72
T1 25.59 24.79 26.20 23.69 25.64
3 Baseline 26.23 23.64 26.83 24.93 27.69
T1 25.30 25.81 23.04 24.00 25.84
4 Baseline 26.00 24.40 25.49 24.31 25.68
T1 24.90 24.66 25.14 24.60 24.67
5 Baseline 25.37 24.87 23.74 22.68 24.46
T1 26.74 26.50 22.94 24.94 24.11
All orgs. Baseline 25.08 24.86 24.95 24.00 25.45
T1 25.52 25.26 24.62 24.21 25.06
Higher scores suggest worsening mental health
Table 40: Trends in the mental health measure by GHQ-12- Minimum score
on scale = 12.
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
+ 0 + 0 0
2
+ - - 0 0 0
3
- +++ - - - - - - - - -
4
- 0 0 0 -
5
++ ++ - ++++ 0
Higher scores suggest worsening mental health
Again the scores are difficult to interpret. Mental health scores could be the
product of a number of different factors. Experiencing negative behaviour and
bullying at work may only represent just one of a number of factors present
both within and outside the context of work.
50
10.7 Organisational outcomes, absenteeism and intention to leave
Absenteeism and turnover behaviour have been the primary focus of previous
research with regards to organisational outcomes associated with bullying.
Although bullying and negative behaviour have been found to be associated
with absenteeism, with the previous BOHRF study indicating a discrepancy in
terms of absenteeism of seven days annually between those who reported
being bullied and those neither bullied nor having witnessed bullying, research
has found a relatively weak link between these variables. Thus, a Norwegian
study reported that bullying only accounted for 1% of total absenteeism
(Einarsen & Raknes, 1992), a figure which increases to 2% in a Finnish study
of heath sector workers (Kivimaki, 2000), and 3% in a recent Swedish study
(Widmark et al., 2005). It therefore does not come as a surprise that it is
impossible to identify any particular trend in the current data.
Table 41: Self-reported absenteeism (time off work last six months)
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 3.87 5.06 5.15 5.84 3.77
T1 1.92 3.60 2.74 1.73 3.13
2 Baseline 2.18 5.21 3.58 2.71 2.74
T1 2.59 5.95 4.59 2.40 4.19
3 Baseline 1.01 3.83 4.42 3.09 3.83
T1 2.26 2.22 5.66 3.15 4.47
4 Baseline 4.55 2.85 2.03 4.37 3.86
T1 10.62 6.86 2.69 3.65 3.68
5 Baseline 5.01 3.25 5.38 5.95 3.96
T1 3.41 2.30 2.96 5.65 2.55
All orgs. Baseline 3.54 4.00 4.19 4.56 3.67
T1 3.72 4.01 3.61 3.25 3.56
A far stronger relationship has repeatedly been found between bullying and
intention to leave, or turnover behaviour (e.g. Unison, 1997; Rayner, 1999;
Hoel & Cooper, 2000).
As the scores in the above table are calculated as a sum of five separate
scores, it would not make sense to represent any changes as a percentage as
done previously in this section with other variables.
51
Table 42: Considered quitting last six months
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline 2.08 2.22 2.30 2.58 2.42
T1 2.19 2.34 2.26 2.52 2.81
2 Baseline 2.77 2.52 2.82 2.36 2.83
T1 2.90 2.76 2.55 2.13 2.32
3 Baseline 3.00 2.38 2.17 2.23 3.06
T1 2.96 2.81 2.63 2.34 2.53
4 Baseline 2.76 2.84 2.73 2.41 2.90
T1 2.60 2.55 2.68 2.55 2.91
5 Baseline 2.13 2.38 2.20 2.39 1.66
T1 2.31 2.53 2.35 2.42 1.82
All orgs. Baseline 2.46 2.46 2.45 2.42 2.49
T1 2.52 2.56 2.47 2.41 2.49
Minimum scale score = 1: Increased scores suggest increased intention to quit
Table 43: Trends in intervention and levels of turnover (data incomplete for
organisation 1 and 5).
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
Al
1
++ + 0 0 +++++
2
+ +++ - - - - - - - - - -
3
0 ++++ +++++
++
+ - - - - -
4
- - - - 0 - 0
5
+++ - - ++ 0 +
Increased scores suggest increased intention to quit
The above figures are difficult to interpret. On the one hand it appears to be
impossible to detect any particular pattern with regard to turnover if one
focuses on relative change. On the other hand, when looking at the post
intervention scores (T1), all units which have received training beyond policy
communication actually report levels of turnover below the organisational
average, which again could indicate a slight change in the right direction.
10.8 Satisfaction with the atmosphere at work
An improvement of the work atmosphere was considered a potential and
desired impact of our interventions. We therefore asked respondents to
consider the following question: “In comparison to six months ago, how
satisfied are you with the atmosphere at work”. Respondents were given three
answer alternatives:’ ‘more’, ‘less’ and ‘about the same’. As can be seen from
the table below, an overwhelming number of respondents considered the
52
work atmosphere to have got worse over the last six months. This might give
some support to the idea that taking part in training may sensitise and make
them more aware of workplace problems. Furthermore, training might raise
expectations which may lead to dissatisfaction when expected change is not
forthcoming.
Table 44: Satisfaction with atmosphere at work compared with six months ago
Intervention group mean scores Organisation Time of
assessment 1 2 3 4 5
1 Baseline -0.02 -0.16 0.17 0.05 -0.25
T1 0.00 -0.14 -0.12 -0.14 -0.35
2 Baseline -0.05 -0.31 -0.03 0.10 -0.10
T1 -0.03 -0.24 -0.36 0.03 -0.04
3 Baseline -0.23 -0.03 -0.22 -0.17 -0.16
T1 -0.04 -0.33 -.046 -0.22 0.11
4 Baseline -0.30 -0.04 -0.55 -0.19 -0.39
T1 -0.24 -0.06 -0.52 -0.24 -0.64
5 Baseline -0.09 -0.15 -0.02 0.12 -0.04
T1 -0.38 -0.24 -.012 -0.27 -0.39
All orgs. Baseline -0.12 -0.13 -0.10 -0.01 -0.18
T1 -014 -0.20 -0.29 -0.17 -0.22
Increased scores suggest improved satisfaction with work atmosphere
As the scale used to measure atmosphere at work goes from –1 to +1 (with 0
representing no change), this can be considered to be a 2-point scale. Thus,
a change in score =0.2 represents a change of 10% (or 0.1 =5%).
Table 45: Trend in ‘satisfaction with atmosphere at work’
Organisation Changes in intervention group mean scores
1
Control
2
Policy
communica.
3
Stress
4
Neg.
Behaviour
Awareness
5
All
1
0 0 - - - - -
2
0 0 - - - 0 0
3
++ - - - - - 0 ++
4
0 0 0 0 - -
5
- - - 0 - - - - - - - -
Increased scores suggest improved satisfaction with work atmosphere
53
10.9 The effectiveness of intervention: assessing ‘objective’ measures
As an alternative measure of the potential efficacy of interventions, we asked
the organisations to provide some ‘objective data’. The response in returning
objective data was varied thus undermining the potential of an appropriate
comparison (see comments in section 7). However, the data was complete for
sickness absenteeism with the exception of one organisation, and for turnover
only two organisations provided complete records at both points of
measurement.
10.10 Intervention and levels of absenteeism
As absenteeism figures varied very substantially from organisation to
organisation and even between units within the same organisation, it makes
little sense to compare raw data or days of absenteeism directly. Instead we
will report on relative change within each intervention group. The three sets of
figures reported thus relate to a) absenteeism level in unit compared to
organisational average prior to intervention (baseline); b) absenteeism level in
unit compared to organisational average post intervention (T1); and c) the
relative change in absenteeism level for unit between baseline and T1 (before
and after intervention). NB: Absenteeism is not reported for organisation 1 as
data for two of the units were reported together and, thus render any
comparison between units impossible. Positive (+) and negative (-) scores
refer to unit levels scores above or below the mean score for the organisation
respectively.
Table 46: Absenteeism comparison in intervention groups
Absenteeism:
intervention group mean scores
Organisation Time of
assessment
1 2 3 4 5
2 Baseline +2.25 +4.52 +4.34 +1.94 +2.44
T1 - 1.70 +1.04 - 1.83 +0.70 - 2.17
Rel. change - 3.95 - 3.48 - 6.17 1.24 4.61
3 Baseline - 0.50 +0.90 +0.60 - 1.00 - 1.40
T1 +4.22 +9.12 +7.52 +7.22 +3.82
Rel. change + 4.72 + 8.22 + 6.92 + 8.22 +5.22
4 Baseline +4.67 +2.02 - 1.21 +0.60 +3.09
T1 +0.24 +1.45 - 0.12 +1.49 - 0.93
Rel. change - 4.43 - 0.57 + 1.09 + 0.89 - 4.02
5 Baseline +2.20 +3.20 - 0.50 +7.20 - 3.45
T1 - 1.50 - 1.15 +2.65 +7.10 - 1.30
Rel. change - 3.70 - 4.35 + 3.15 - 0.10 2.15
Closer scrutiny of the above figures reveals no clear patterns with regard to
change in absenteeism associated with any particular intervention.
As has been argued previously in this report, whilst the association between
bullying and negative behaviour, on the one side, and, absenteeism on the
other, is weak, a much stronger association is found for intention to leave.
54
Table 47: Intervention and levels of turnover (data incomplete for
organisation 1 and 5).
Turnover:
intervention group mean scores
Organisation Time of
assessment
1 2 3 4 5
2 Baseline
T1
Rel. change
3 Baseline - 2.80 - 0.30 - 1.30 +4.40 +0.10
T1 - 1.20 - 3.30 - 2.00 - 3.10 -2.30
Rel. change + 1.60 - 3.00 - 0.70 - 7.50 - 2.40
4 Baseline +0.61 +3.31 - 0.73 - 4.40 - 3.92
T1 - 0.76 +7.06 - 1.66 - 5.44 - 1.64
Rel. change - 1.37 + 3.75 - 0.93 - 1.04 + 2.28
Unfortunately, the data for turnover is difficult to interpret. Whilst most
experiments groups show a reduction in turnover, this trend is not consistent
across all experiment groups.
10.11 Other ‘objective’ measures
Only one organisation provided all requested data. However, in this case no
changes were recorded between baseline and post-intervention measurement
(T1) for any of the requested categories.
10.12 Total assessment of effectiveness of interventions
The previous sections have revealed that it is very difficult to trace particular
trends in the data when looking at different variables which could have been
influenced by the interventions. To enable us to consider all key variables
together with the aim of looking for potential trends across the data, we put all
the ‘relative representations’ by means of increases or decreases in variables
scores into one table (see table 48 below).
55
Table 48: Assessment of interventions (desired direction of change for each
variable is given in brackets)
O
R
G
IN
T
BU
(-)
WB
(-)
NAQ
(-)
GHQ
(-)
PC (+) ATM
(+)
TO
(-)
1 1 0 - 0 + 0 0 ++
2 1 + 0 - - + - 0 +
3 1 - 0 + - - + + 0
4 1 0 + + + 0 - 0 0 -
5 1 - - 0 + + - 0 + + +
1 2 0 + - - 0 0 0 +
2 2 - 0 - - - - - 0 0 + + +
3 2 0 0 + + + + - - - - +++ +
4 2 - - 0 0 - - 0 - - -
5 2 + 0 0 + + 0 0 - -
1 3 0 0 ++++++ + 0 - - - 0
2 3 + + - +++++++++ 0 - - - - - - -
3 3 0 - ++ - - - - - 0 - - +++ +
+ + +
4 3 - 0 - - - - 0 + 0 0
5 3 0 - - - - 0 - + +
1 4 - + 0 0 + - 0
2 4 - - ++++++ 0 0 0 - - -
3 4 - - - - - - ++ 0 +
4 4 +++ - - 0 0 - - 0 -
5 4 0 + 0 ++++ - - - - - 0
1 5 0 +++ ++ 0 - - +++ +
+
2 5 - - - - - - - - - - - - - - - - - - 0 0 0 - - - - -
3 5 0 0 - - - - - - - - - - - +++ + + - - - - -
4 5 + 0 ++++++ - +++++ - - 0
5 5 0 0 ++ + 0 - - - +
(BU=Bullying; WB=Witnessed bullying; NAQ=Negative behaviour;
GHQ=mental health; PC=Psychological contract; ATM=Atmosphere;
TO=turnover/decision to leave).
For nine (45%) of the 20 experiment groups (table 49), the changes are in the
desired direction for most variables, over and above those observed in the
control group. In all of these cases there were improvements with regard to
negative behaviour and bullying (which was the key objective of the
interventions) and an improvement (or at worst no change) with respect to the
56
other variables measured (i.e. all variables with the exception of atmosphere
and turnover, both of which were measured by one-item measures).
For three of the experiment groups (highlighted in darker grey) this trend was
upheld across all variables with very considerable improvement reported for a
number of them. It is important to notice that two of the groups represented
units that had received a full days training or taken part in all three workshops,
whilst the third group had received negative behaviour awareness training.
Although great care must be taken when interpreting these findings, they do
appear to indicate that the training might have had the intended effect, in at
least some cases.
11. Intervention feedback from the trainer and participants
11.1 The trainer’s perspective
As part of the evaluation process, the trainer was asked to report back on
each scheduled session on a specifically devised form. In addition to receiving
written feedback, the researchers conducted a two-hour interview with the
trainer to explore some of his perceptions around the impact of the
intervention programme in each of the five organisations. In total the trainer
carried out 40 distinct interventions.
The trainer reported varying degree of success. In particular, the training
groups appeared to vary with respect to group-dynamics (sometimes cross-
questioning rather than interacting) and their ability to engage with the training
content/message. In this respect the climate of the workshop would impact on
process, what may be discussed and the outcomes.
Looking at the three different workshops the trainer made the following
comments:
Policy communication:
Were only effective when a large number of relatively senior managers
attended the session.
Stress management:
The nature of the workshop meant that they often become too focused on
‘blaming the individual’. Furthermore, problems rather than solutions often
took centre stage when organisational issues were emphasised. The
workshops were often undermined by the fact that they focused on
organisational issues which participants felt they were unable to influence due
to lack of power/ seniority. However, there appears to be a consensus that a
lot of negative behaviour could be removed by reducing stress levels.
Negative behaviour awareness:
As most people do not admit to experiencing bullying behaviour, time is
needed for problem recognition to take place. The presence of macho
attitudes or fear of not being politically correct may undermine the ability to
learn. In general, the training was more effective in dealing with negative
57
behaviour than raising awareness of one’s own role in creating or contributing
to the problem. The transactional models introduced to deal with the problem
were considered effective only when participants were able to relate their own
experience directly to the model.
The following issues were highlighted by the trainer:
• The establishment of constructive dynamics and sufficient time is
needed for experiential learning to take place
• With regard to quantity and quality of training, a minimum of one day
appeared to be most effective. This is also evident from some of our
questionnaire findings, which suggests that individuals who attended
the full day session covering policy communication, stress
management and negative behaviour awareness training, gained the
most out of the interventions.
• One of the main objectives of this research was to implement bespoke
interventions that were developed with the benefit of local knowledge
and which provided examples that were relevant to the target audience.
Thus, for the learning to be effective the audience had to be able to
relate to the models presented.
• Groups need to be homogenous in that individuals attending sessions
should be from similar levels of the organisation to create a more
secure environment.
• Key people were not invited to training sessions or they did not attend
when invited: the training would be more effective with people from a
position of power or who were more able to influence their
circumstances.
• Several external factors may influence the effectiveness of the training
including substandard rooms, noisy environments and busy
organisations with frequent interruptions / individuals leaving
prematurely to attend to their work.
11.2 Feedback from participants
Participants were provided with a one-page evaluation form to complete at the
end of all stress management and negative behaviour awareness training
sessions. A total of 193 evaluation forms were collected from all five
organisations.
Three quarters of respondents responded positively in response to whether
the training challenged individuals to think in new ways.
58
Individuals also responded very positively to ‘How relevant was the session to
the programme aims’, ‘How interesting did you find the session’ and ‘How
would you rate the overall content of the session’ by rating each item as
‘good’ to ‘very good’ (1 =poor, 5 =very good).
In addition to this, participants were asked to write down the main thing they
had learnt from coming to the workshops. Their responses include:
• Self appraisal / reflection
• Awareness of different types of bullying behaviour
• To ensure that communication channels are kept open
• To control my stress levels better
• Prioritise
12. Post Intervention focus groups
Altogether eight post-intervention focus groups were carried out in three
organisations.
The aim of these focus groups was to qualitatively assess the effectiveness of
the interventions. For that reason the focus group discussions concentrated
on issues such as a) what are the key issues you remember from your
Di d the trai ni ng chal l enge you to
thi nk i n new and di fferent ways?
Yes:
75.10%
No:
24.90%
4.33
4.23
4.27
1
1.5
2
2.5
3
3.5
4
4.5
Relevance Interest Overall
rating
Mean participant
feedback scores
59
training and b) to what extent has the training changed your own or others’
thinking and behaviour with regards to bullying and negative behaviour.
The discussions indicate that specific training content is still remembered and
considered making sense six months after it was received. The following
issues were emphasised in the discussions:
“I think as managers we’re all aware of it, we’ve been on diversity
training, but that workshop was quite good in defining the different
aspects of bullying” (female manager, Civil Service).
“I think it reinforced for me how to look at different ways of managing
conflict in strained environments. For me it was a good tool to enhance
what I already learnt as a trainer…you’re just standing back and
allowing everybody to express themselves in whatever way is good for
them and then almost offering a mini-counselling service before you
get on with what you’re trying to achieve and that has certainly helped
me build a relationship with the people I work with” (female trainer,
Police Force).
In particular a number of participants highlighted the usefulness of the
transactional analysis approach which formed part of the Negative Awareness
workshops:
“The different models…child / adult relationships….for a piece of
information it was useful and it’s nice as an observing thing to say “oh
right, you’re behaving like a child and I’m in parent mode and we’re not
quite negotiating”…It allowed me to take one step back and reflect”
(male Officer, Police Force).
“It gave us an insight and awareness. You go back and see things in a
different light. I think it was the transactional analysis that you can talk
to someone as a parent and they can respond as a child and you need
to be on an adult to adult level all the time… I thought that was good”
(female administrator, NHS).
“I’ve used it (transactional analysis). I’m not saying how I’ve used it has
been successful in reflecting on my own performance. I’ve used it to try
and get a different outcome and sometimes it’s worked and sometimes
it hasn’t. You’ve got to be aware of the recipient of your message…just
may not want to know” (male Officer, Police Force).
Some people emphasised that the training had particularly helped them
improving their understanding of the problem:
Defining the problem
“I think as managers we’re all aware of it, we’ve been on diversity
training, but that workshop was quite good in defining the different
aspects of bullying” (female manager, Civil Service).
60
“I came away with was that it was to inform staff, colleagues etc of the
situation so that it was not considered as an overt activity, something
that was going on regularly…sort of making sure that those negative
traits and tendencies were managed better because some of them
were to improve productivity or certainly [make people] more motivated
with what they were doing. They needed to modify their approach so
that people weren’t put under undue pressure” (male, Civil Service).
Impact on behaviour
“I react differently to aggression now. Aggression is just self expression.
It may not be necessarily a personal attack on you but a lot of people
tend to deal with aggression with aggression. But I take myself out of
that and going in to that session reinforced what I need to and put into
place for my own protection before I actually react to a situation”
(female trainer, Police Force).
“It had made me aware that behaviours that I had seen could be
perceived as bullying, that a situation that I was managing someone I
needed to be mindful that my approach to a subordinate would need to
be necessarily open in ways which would project me coming into a
situation where someone can say “you’re bullying me”. No. Leaving the
door open at all points for someone to express, you know if they’ve got
reservations, and doing it in a manner if they did that they would feel
assured that we can find a way round that” (male, Civil Service).
“I think what we’re talking about is more broader than bullying and
harassment, and we’re talking about negative behaviours that we have
observed or experienced which of themselves may not be as painful as
individual instances of bullying but are equally relevant to the
department in terms of getting a happier more productive workforce.
And therefore to focus exclusively on bullying or harassment isn’t
terribly productive because I might sit there and think I wouldn’t bully
anybody or harass anybody and I have never had experience of those
things but I’ve had plenty experience of less painful negative
behaviours which I myself would like to learn from and would
encourage others to learn from. And I think this sort of drip feed will
then help them to reduce instances of bullying and harassment” (male,
Civil Service).
“It certainly is there in my mind because if the training that this is not
punitive action…I think that the one thing that the training will have
done is it will have usefully indicated to people that the perspective that
they will view their behaviour at with subordinates, with colleagues,
superiors etc, that’s something that needs to be analysed. It’s maybe
the hardest thing as individuals that we face is to examine ourselves as
best as we can from the outside…” (male, Civil Service).
61
Indirect suggestions that behaviour has changed due to interventions
(individuals more aware of their own behaviour)
“The Force is taking great steps to change things. The danger sometimes is
that it perhaps has gone too far…you’re very mindful that you’re not
bullying…There is a fine line between the definition of bullying and actually
having to tell someone that that needs to be done” (male Officer, Police
Force).
“we joke and muck around as friends would but at the same time we are
slightly uncomfortable with doing so because we feel quite heavy on our
shoulders, the responsibility to be diverse and not to intimidate anyone or
harass them or bully them and actually that can be negative because it
inhibits what in other circumstances would be normal interaction and so I find
myself saying “hold on, you shouldn’t be saying that sort of thing” (female,
Civil Service).
Uncertainty or suggestions that interventions may not have worked or
behaviour
“It’s a long haul process. I had a situation not so long ago where the
nurse actually turned around and said she didn’t want to take it further
because she felt that it wouldn’t get anywhere… if people feel like that
then it isn’t working very well” (female nurse manager, NHS)
“Well if people obviously have a problem then they will now know
where to go…the information is there if you need it, but I don’t think the
overall perception of the attitude of the management board has
changed…things have changed at the grass roots level” (female, Civil
Service)
“I think the Force has worked on it over the years. It certainly has not
eradicated it…I think the Force is putting forward a positive message
that bullying will not be accepted or tolerated. That has potentially had
the effect of making everybody know where we stand and perhaps
putting people off what may be perceived as bullying behaviour but on
the other hand it also has the effect of driving it underground and
making bullies more devious…I think people use less obvious ways of
wielding their power or influencing other people because they know
they face potential discipline if they get caught” (male Officer, Police
Force).
“Because I haven’t had any major issues within my team setting, we
haven’t had to test the processes so I don’t really know” (female nurse
manager, NHS).
“Impossible to say, I’ve changed jobs, different people.” (Male, Civil
Service).
62
What else can the organisation do?
Several participants emphasised the importance of ongoing initiatives and
training in order to bring about change.
“It was all relevant. Just having one session it’s difficult to enable it to
be a part of your life…It’s got to be on an on-going basis then it
strengthens you as a character to be able to deal with any conflict and
manage yourself better…It’s got to be on-going” (female trainer Police).
“…the training has had a vague but definitely instilling influence in me
and my approach” (male, Civil Service).
However, the answer is not necessarily associated with training per se but
with who receive the training. In other words, to be effective the training needs
to be targeted at those in need of training.
“We feel that we are…lower down the ladder when it comes to the
supervisory ladder, we think it should have gone up higher.” (female
administrator, Police Force).
“I suppose it comes down to how people are selected for training in the
first place” (female nurse, NHS).
“I think part of the problem is that you are preaching to the converted. I
don’t think we would be here if we were not aware of the bullying and
harassment as an issue anyway” (male, Civil Service).
“You’ve got a focus group of five people who’ve put themselves
forward. Probably people slightly on the goody side of normal, probably
the people that don’t need this {training]. The people who you do need
here are the people who didn’t put their hands up and aren’t here.”
(male, Civil Service).
From the focus groups discussions referred to above we can conclude that
although the training receives good marks and for some is considered to have
led to behavioural change, the picture is somewhat mixed.
63
13. Discussion
The present study confirms previous UK findings, suggesting that workplace
bullying is a serious workplace problem affecting a substantial number of the
workforce either directly or indirectly. The inclusion of some research
instruments or scales used in the nationwide study sponsored by BOHRF and
published in 2000 (Hoel & Cooper, 2000), has also made it possible to make
some comparisons with previous findings. Although an overall rate of bullying
of 13.6 and 14.3% at the two measurement times is somewhat higher than the
national average of 10.6% emerging from the study in 2000, most of the
current findings largely replicate those of the nationwide study and, thus give
strong support to the validity / authority of the larger scale study. Moreover,
with levels of bullying varying from 10.8 to 23.0% in the five organisations
taking part in the study, the identification and implementation of interventions
which could contribute to rectify the problem appears to be entirely justified as
well as timely.
As highlighted in the report, a risk-assessment instrument has successfully
been developed and tested. The validity of this instrument, the 29-item
Bullying Risk Assessment Tool (BRAT) was confirmed, with the BRAT (29
items) and each of its five factors, correlating highly with negative behaviour
and bullying as well as with mental health. Moreover, each of the five factors
independently predicted negative behaviour, whilst all factors with the
exception of ‘workload’ predicted self-labelled bullying. Although it is too early
to anticipate its full practical value, it is hoped that it will be tested out in other
studies in the near future. The BRAT represents the first attempt anywhere to
apply a risk-assessment approach to the issue of workplace bullying and as
such, must be considered a milestone. It might not yet have found its final
form but international interest in replicating or testing its qualities, particularly
from research teams in Belgium and Denmark, speaks for itself.
The study’s primary aim was to develop, implement and evaluate
interventions which alone or in combination could prove to be effective in
reducing negative behaviour at work and ultimately preventing workplace
bullying. To achieve this aim a highly complex and scientifically rigorous study
was designed and carried out. By involving five organisations from four very
different environments, with their own very specific structures, cultures and
routines, the researchers have had to juggle a multitude of issues in order to
carry out the study according to the original protocol and to ensure that
sufficient information was collated at various points. With this in mind we can
conclude that we have reached our overall aims and objectives.
The interventions applied in the study were all based on sound theory and
current understanding of bullying and its antecedents, and we believe that
they have adequately addressed the problem in order to bring about the
envisaged behavioural change. Similarly, the professionalism of the trainer
and the effectiveness with which the workshops were delivered were never in
question, with feedback from participant also appearing to give strong support
to this. However, with reference to the stress literature, locally developed and,
64
thus, fully contextualised training programmes developed in response to
particular local needs could be anticipated to have an even greater impact.
But the scientific design which aimed to compare interventions across
different settings, and thus allow for conclusions with regards to
generalisability and transferability of findings to be made militated against
such an approach.
Unfortunately, we have only to a limited extent been able to utilise the
objective measure which were obtained during the study. A number of factors
accounts for this shortcoming: Firstly; some organisations were unable to
obtain the data as requested (and agreed) due to difficulties with external data
processing mechanisms and incompatible formats where intervention units
were different from internal units of measurement. Secondly; organisations
were unable to obtain information about particular issues, e.g. grievance and
complaints records. With hindsight, such shortcomings could possibly have
been resolved by involving the organisations in developing the actual
measures as well as testing the system prior to the study.
Despite strong indications from participants that the training was relevant and
at least to some extent has had the intended behavioural impact as witnessed
by focus group accounts, there is insufficient evidence in the data to make
any conclusions with regard to the efficacy of particular interventions. This is
unfortunate, albeit not surprising, and seems to be in line with most
organisational intervention research (Murphy & Sauter, 2003). However, when
comparing the data across a number of variables, it is apparent that some
improvement in the desired direction for negative behaviour and bullying has
taken place in approximately a third of experiment groups, but with different
interventions associated with positive outcomes in different organisations. If
we consider these results with the feedback from participants on the day of
training and six months following the training, the trainer’s reports on delivery
and anticipated effectiveness, as well as focus group participant observations
and responses post intervention, we can tentatively conclude that the training
does seem to have at least some effect, although different interventions may
work better within different contexts. Moreover, for two of the experiment
groups all scores on the eight key variables were in the desired direction, with
a very substantial improvement noted for some of the variables. In both cases
these results have emerged from units where the experiment group have
received a full day’s training. The fact that this finding was not upheld across
all five organisations can be due to a variety of reasons, with the trainer’s
report in one particular case suggesting considerable unwillingness/resistance
on behalf of participants to engage with the training, and another,
experiencing threat of redundancies at the time of post-intervention
measurement.
65
To account for the overall results a number of factors need to be considered:
Intervention may increase awareness and expectations, and in
turn lead to dissatisfaction, affecting post-intervention
measurements.
It should be emphasised that raised levels of problem awareness
during the initial stages in a change process should not be considered
a problem per se and may even represent an asset. However, it is well
known from other studies (e.g. stress literature) that interventions might
contribute to sensitisation of a problem with higher scores on particular
variables as a result. In a similar way, interventions may also raise
expectations which subsequently could manifest itself in dissatisfaction
when change does not occur or when the pace of change is considered
to be too slow. For example, it is possible that by identifying and raising
awareness about stressors present in the work environment but at the
same time failing to, or being unable to provide participants with
necessary powers to address these stressors, further stress and
frustration may emanate. Following this line of argument it might,
therefore, be suggested that where no change has occurred between
the two measurement points, an actual improvement could have taken
place. The eight percentage point drop in response rate from
measurement one to measurement two might also has impacted on
results in a similar way, reinforcing the above argument. Thus, it is well
known from previous studies (e.g. Einarsen et al., 1994) that targets of
bullying are somewhat more likely (Einarsen and colleagues found a
difference of 10%) to respond that those who did not consider
themselves to be bullied.
The right people may not have been trained
As previously outlined, in order to have the maximum effect, our
intention was to aim the training at people in line management
positions. Although there is no problem in including employees without
supervisory responsibility in the training, where this has lead to line
managers being replaced by rank and file employees, it might have
reduced the maximum potential impact of the training.
There is some evidence to suggest that individuals who were
considered (by their managers or their colleagues) as potential
beneficiaries of the training, were actually not selected for training.
Moreover, there is also anecdotal evidence that some actually
deliberately avoided taking part in the training despite being selected,
by not turning up at the session or by nominating colleagues in their
place. Insufficient attention to planning and scheduling on behalf of the
participating organisations has in some instances resulted in low
participation rates and in other cases, people have left in the middle of
a focus group or a training session because of other work
66
commitments. Even with forewarning, participating individuals &
organisations have had to come to a decision as to whether to proceed
with crucial day-to-day work tasks, such as operating on patients, or to
attend focus groups / training sessions. It should be noted that
resource limitations on behalf of the researchers have meant that the
interventions have had to been planned on two consecutive days within
each organisation. Although this has worked well for research
purposes, in has caused organisational problems and possibly
contributed to lower participation rates. In principle the training could
have been split into different modules and could be delivered more
flexibly over a longer period of time.
By contrast, there is evidence to suggest that management have
consciously or subconsciously made a decision to apply training to
particular ‘hot spots’ or problem areas. Whilst this might contribute to
inflating bullying figures and other measurements, suggesting that the
data might not be representative of the entire organisation, such
decisions make complete sense and should in principle not affect the
analysis where the emphasis has been on relative change between
post-intervention (T1) and baseline measures.
Things are going on within the organisations impacting on
processes and individual outcomes
The post-intervention focus groups bear witness of the presence of
other factors potentially influencing findings. For example, in some
cases other initiatives aimed at organisational change, of which some
were also associated with training of participants, were going on at the
same time or in parallel with our project. Moreover, other organisational
processes may interfere with or even counteract our effort to bring
about change, e.g. restructuring or organisational change. For example,
in one of our organisations a reorganisation unknown to us at the offset
of the project, not only led to internal turmoil but also the potential of a
large number of redundancies.
The units of comparison (referred to as intervention group 1-5)
may not have remained the same during the entire course of the
study.
There is evidence that some individuals have been transferred to other
jobs or other units or may have left the organisation altogether. Others
may have been recruited during the period following the intervention.
However, this is to be expected and is likely to have only marginally
affected the results. Furthermore, in at least one of the organisations a
large number of employees from one of the experimental units were
moved to another unit in the middle of the study. Taking turnover rates
into account (approximately 10 percent), such factors would undermine
the value of straightforward comparison of pre and post intervention
measures.
67
Overall ‘critical-mass’ may not have been achieved.
Although we were up against the reality of limited resources and
commitment of the participating organisations, it may be argued that
too few people within each unit were trained in order to have any
measurable effect in the desired direction. Similarly, the amount of
training each individual received may have been insufficient for any
behavioural change to occur. In this case both factors are likely to have
influenced outcomes. The fact that participants selected for training
frequently did not turn up or, in some cases had to leave in the middle
of the training further highlight this problem. However, as emphasised
in the previous point, ‘critical mass’ may be as much related to the
total number of people trained as well as the amount of training
participants received.
The time between interventions and post-intervention
measurement (T1) is too short for any effects to have occurred.
For most intervention studies effects appear to be the greatest
immediately after the training where the content is relatively fresh in
the minds of participants, and then gradually diminishes over time,
especially where the training is not repeated (Giga, Cooper, Faragher,
2003). However, the nature of the problem of destructive behaviour
and bullying and its association with the culture of the unit and the
wider organisation could suggest that change would be slow and
possibly a delayed process. However, the relatively short period
between measurements could not to be avoided due to the total
duration of the study and the expressed wishes of the participating
organisations. To test this assumption a second measurement (T2), i.e.
12 months after the intervention is currently being discussed with one
of the organisations. It is questionable whether a T2 measurement
would necessarily bring about greater clarity bearing in mind some of
the problems discussed above, which may to some extent have had an
affect on this organisation.
The above discussion reviews a number of factors which alone or in
combination may help to explain why it was not possible to trace any clear
pattern in the quantitative data and, thus, prevent us from drawing any firm
conclusions with regard to the effectiveness of any particular intervention or
combination of interventions. However, looking at the complete evidence
presented, the combination of quantitative and qualitative evidence appears to
provide some evidence that the intervention may have had a small positive
effect, particularly for those attending the full day training.
Whilst it might be somewhat disappointing not to be able to draw any clear
conclusions with regard to the efficacy of applied interventions, this study, the
first of its kind with respect to bullying and destructive behaviour, has provided
us with considerable insight into organisational intervention processes, and
which therefore is likely to greatly benefit future studies. It is to such learning
points we now turn our attention.
68
14. Conclusions
As far as we are aware, this report completes the very first scientific study into
evaluating the effectiveness of management interventions for tackling
interpersonal conflict in the workplace. Despite many obstacles and
challenges on the way, we are confident that we have met our original aims
and objectives in connection with this ambitious study. A risk assessment tool
(BRAT) to assess the risk of negative behaviour and bullying has been
successfully developed and validated. In our opinion, the BRAT has
considerable scope for practical application.
By means of a complex study design, and by applying rigorous scientific
standards throughout, we developed, implemented and evaluated three
different interventions, all theoretically sound and based on a review of the
literature and local contextual information. In order to assess the effectiveness
of the interventions or combinations of interventions, we applied a randomised
controlled trial (RCT) design and used a wide range of data of a quantitative
as well as quantitative nature in the validation/evaluation process. Five large
organisations, within the public sector took part in the study and the same
design was adopted for all organisations to allow for evaluation purposes and
generalisation of findings.
With regard to the survey data obtained prior to, and six months after the
interventions, no statistical differences emerged for any key variable between
intervention groups across the five organisations. Thus, it was impossible to
establish the efficacy of particular interventions or, combinations of
interventions. However, when all outcome variables were considered together,
an improvement or change in the desired direction appears to have taken
place in some of the intervention groups, particularly where participants had
taken part in a full day’s training and, thus, received the complete training
programme. These findings were also supported to some degree by feedback
of a qualitative and quantitative nature from the trainer, trainees and
participants of post-intervention focus groups, whether they had previously
had taken part in the training or not. When the same outcomes were not
reported across the five organisations, various factors, some of which were of
an organisational nature and could be accounted for, might have contributed
to this result. Firstly, although it was our intension to primarily focus the
intervention on managers or people with supervisory responsibility in order to
achieve the maximum impact with limited resources, there is clear evidence
that in many cases this was not the case, with employees without supervisory
responsibility, replacing members of the key target groups or, indeed,
individual managers or supervisors identified as in need of training. Secondly,
and related to the previous point, with only 5-20% of staff within a
measurement unit being trained, insufficient numbers or a ‘critical mass’ was
not achieved to bring about a measurable effect. Thirdly, other factors of an
organisational nature or the amount of training provided may have been
insufficient for experiential learning to take place. Fourthly, it is well known
from other intervention studies that training or other interventions might
sensitise participants and the environment within which the interventions are
69
taking place to the problems, with greater awareness manifesting itself in the
shorter term in dissatisfaction and inflated scores on certain variables.
With regard to the apparent success of the full day’s training combining all
three interventions, this could have as much to do with the amount of training
as the particular nature of the training. Thus, we will second the view of the
trainer that in order to achieve the desired change in behaviour, experiential
learning must take place. For a sensitive issue such as the one in question,
this would require positive group dynamics and the development of a safe
learning environment, both of which in most cases take time to establish.
The present study has provided us with many important learning points,
particularly with regard to the intervention process which we believe will
greatly benefit future intervention studies in the area of workplace bullying,
destructive behaviour and beyond. In this respect we would emphasise the
importance of continuous management commitment to the process, based on
shared in-depth understanding of the aims of the interventions, responsibility
for resource provision, resource allocation and, implementation of the entire
research project. However, to ensure a successful outcome wide participation
and involvement must be established from onset, allowing for interventions to
be fully contextualised or anchored in the local context. In this respect it is
tempting to suggest that the scientific rigour and RCT design applied to the
present study might in itself has militated against a better or more clear-cut
result. Thus, we have previously been warned that applying rigorous scientific
design when undertaking research in busy, rapidly changing environments is
fraught with difficulties (Kompier, 1999). Therefore, whilst a case-study design
where the interventions are designed to address the needs of the organisation
and based on wide employee involvement would make it more difficult to
make predictions with regard to generalisability and, indeed, transferability of
findings or tested approach, a case study approach might have been more
successful in establishing the efficacy of interventions.
Finally, our results suggest that there appears to be no ‘quick fix’ solution to
alleviate an organisation of the problems of destructive behaviour and bullying
or to ensure that bullying does not manifest in the future. In our opinion, the
only viable option open to the organisation is to embark upon cultural change,
a process that can be long and difficult and which would require continuous
management commitment. But as many organisations come up against
restricted resources despite their best intentions, it is important not to judge all
interventions within a limited scope or timeframe as a dissipation of resources,
or worse, as counterproductive. Thus, carefully planned and targeted
interventions based on sound theory could bring about change and may
represent a first important step on the ladder of cultural change, even though
definite evidence of any desired impact might be difficult to measure,
particularly in the short-term.
70
15. References
Appelberg, K., Romanov, K., Honlasalo, M., & Koskenvuo, M. (1991)
Interpersonal conflicts at work and psychosocial characteristics of employees.
Social Science Medicine, 32, 1051-1056.
Ashforth, B. (1984) Petty tyranny in organizations. Human Relations, 47 (7),
755-778.
Bellenger, D.N., Bernhardt, K.L. & Goldstucker, J .L. (1976) Qualitative
Research Technique: Focus group interviews. Qualitative Research in
Marketing. Chicago: American Marketing Association.
Clarke, S. & Cooper, C.L. (2004) Managing the Risk of Workplace Stress.
London: Routledge
Cooper, C.L. & Cartwright, S. (1997) An intervention strategy for workplace
stress. Journal of Psychosomatic Research, 43, 7-16.
Cox, T., Griffith, A. & Rial-Gonzalez, E. (2000) Research on Work-related
stress. Luxembourg: European Agency for Safety and Health at Work.
Einarsen, S. and Hoel, H. (2001) The validity and development of the Revised
Negative Acts Questionnaire. Paper presented to the 10
th
European Congress
of Work and Organizational Psychology, Prague, 16-19 May.
Einarsen, S., & Raknes, B. I. (1991). Mobbing i arbeidslivet. En undersokelse
av forekomst og helsemessige konseqvenser av mobbing pa norske
arbeidsplasser. (Bullying at work: a survey of frequencies and health
consequences of bullying at Norwegian workplaces). Forskningssenter for
Arbeidsmilijø: Universitetet Bergen, Norway
Einarsen, S., Hoel, H., Zapf, D., & Cooper, C. L. (Eds.) (2003). Bullying and
Emotional Abuse in the Workplace. International perspectives in research and
practice. London/New York: Taylor and Francis.
Eisenberger, R., Huntingdon, R., Hutchinson, S. & Sowa, D. (1986) Perceived
Organizational Support. Journal of Applied Psychology, 71(3): 500-507.
Giga, S. I., Cooper, C. L. & Faragher, B. (2003) The Development of a
Framework for a Comprehensive Approach to Stress Management
Interventions at Work, International Journal of Stress Management. Special
Issue: Stress and Its Management in Occupational Settings, 10(4): 280-296.
Goldberg & Williams (1988) A User’s Guide to the General Health
Questionnaire. Windsor: NFER-NELSON
71
Hoel, H. & Cooper, C.L. (2000) Destructive Conflict and Bullying at Work,
Manchester School of Management, University of Manchester Institute of
Science and Technology
Hoel, H. & Salin, D. (2003) Organisational antecedents of workplace bullying.
In S.Einarsen, H.Hoel, D.Zapf & C.L.Cooper (Eds.) Bullying and Emotional
Abuse in the Workplace. International perspectives in research and practice
(203-218). London/New York: Taylor and Francis.
Hoel, H., Cooper, C. L., & Faragher, B. (2001). The experience of bullying in
Great Britain: The impact of organisational status. European Journal of Work
and Organizational Psychology, 10, 443-465.
Hoel, H., Faragher, B. and Cooper, C.L. (2004) Bullying is detrimental to
health, but all bullying behaviours are not equally damaging. British Journal of
Guidance and Counselling, 32, 3, 367-387.
Hubert, A.B. (2003) To prevent and overcome undesirable interaction: A
systematic approach model. In S. Einarsen, H. Hoel, D. Zapf & C.L.Cooper
(Eds.) Bullying and Emotional Abuse in the Workplace. International
perspectives in research and practice (299-311). London/New York: Taylor
and Francis.
Kitzinger, J . (1994) The methodology of focus groups: The importance of
interaction between research participants. Sociology of Health and Illness, 16,
(1), 103-121.
Kivimäki, M., Elovainio, M., & Vahtera., J . (2000) Workplace bullying and
sickness absence in hospital staff. Occupational and Environmental Medicine,
57, 656-660.
Knorz, C. & Zapf, D. (1996) Mobbing – eine extreme Form sozialer Stressoren
am Arbeitsplatz. Zeitschrift für Arbeits- & Organisationspsychologie, 40, 12-21
Kompier, M., Geurts, S., Grundemann, R., Vink, P., & Mulders, P. (1998)
Cases in stress prevention: The success of a participative and stepwise
approach. Stress Medicine, 14, 155-168.
Kräkel, M. (1997) Rent-seeking in Organisationen - eine ökonomische
Analyse sozial schädlichen Verhaltens. Schmalenbachs Zeitschrift für
Betriebswirtschaftliche Forschung, 49 (6), 535-555.
Liefooghe, A.P.D., & Mackenzie Davey, K., 2001, Accounts of workplace
bullying: the role of the organization. European Journal of Work and
Orgniazational Psychology, 10 (4), 375-392.
Liefooghe A.P.D. and Olafsson, R. (1999) ‘”Scientists” and “amateurs”:
Mapping the bullying domain.’ International Journal of Manpower, 20, 1&2,
39-49.
72
Littler, C.R., Bramle, T. & MacDonald, J . (1994) Organisational restructuring:
downsizing, delayering and managing change at work. Industrial Relations
Research Series, No. 15, AGPS, The Department of Industrial Relations,
Canberra, December.
Murphy, L.R. and Sauter, S.L. Work Organization Interventions: State of
Knowledge and Future Directions. Cincinnati (OH), National Institute for
Occupational Safety and Health.
Neuman, J .H. & Baron, R.A. (1996) Workplace violence and workplace
aggression: evidence concerning specific forms, potential causes, and
preferred targets. Journal of Management, 24, 391-419.
Quinlan, M. (1999) The implications of labour market restructuring in industrial
societies for occupational health and safety. Economic and Industrial
Democracy, 20, 427-460.
Rayner, C. (1999). From research to implementation: Finding leverage for
prevention. International Journal of Manpower, 20, 28-38.
Rayner, C., Hoel, H. & Cooper, C.L. (2002) Workplace Bullying: What We
Know, Who is to Blame, and What Can We Do? London: Taylor and Francis
Robinson, S. L. (1996) Trust and Breach of the Psychological Contract.
Administrative Science Quarterly, 41: 574-599.
Rousseau, D.M. (1995) Psychological Contracts in Organisations:
understanding written and unwritten agreements. Thousand Oaks: Sage.
Salin, D. (2004) Ways of explaining workplace bullying; a review of enabling,
motivating and precipitating structures and processes in the work environment.
Human Relations, 56.
Lewis, D. & Sheehan, M. (2003) Workplace bullying: theoretical and pratical
approaches to a management challenge. International Journal of Decision
Making, 4: 1-10.
Sparks, K. & Cooper, C.L. (1999) Occupational differences in the work-strain
relationship: Towards the use of situation-specific models. Journal of
Occupational and Organizational Psychology, 71, 219-229.
Spurgeon, A. (2003) Bullying from a risk management perspective. . In
S.Einarsen, H.Hoel, D.Zapf & C.L.Cooper (Eds.) Bullying and Emotional
Abuse in the Workplace. International perspectives in research and practice.
(327-338). London/New York: Taylor and Francis.
Stewart, M.B. and Swaffield, J .K. (1997) ‘Constraints on the desired hours of
work of British men. Economic Journal, 107, 520-535.
Turner, B.A. (1978) Man-made Disasters. London: Wykeham.
73
UNISON. (1997). UNISON members experience of bullying at work. London:
UNISON.
UNISON (2000) Police Staff Bullying Report (number 1777), London:
UNISON
Vartia, M. (1996) The sources of bullying: Psychological work environment
and organisational climate. European Journal of Work and Organizational
Psychology, 5, 203-214.
Wharton (1992) Risk mmanagement: basic concepts and general principles.
In J . Ansell and F. Wharton (eds.). Risk Analysis, Assessment and
Management, Chichester: J ohn Wiley.
Zapf, D., Einarsen, S., Hoel, H., & Vartia, M. (2003). Empirical findings on
bullying. In S. Einarsen, H. Hoel, D. Zapf, & C. L. Cooper (Eds.), Bullying and
Emotional Abuse in the Workplace. International perspectives in research and
practice (103-126). London: Taylor and Francis.
Zapf, D., Knorz, C., & Kulla, M. (1996). On the relationship between mobbing
factors, and job content, the social work environment and health outcomes.
European Journal of Work and Organizational Psychology, 5, 215-237.
74
16. Appendices
Appendix A: Advisory Board Membership
Baroness Anne Gibson Chair of Committee,
Member of the House of Lords.
Rob Allan Director of Human Resources,
West Hertfordshire NHS Trust.
Janet Asherson Head of Environment, Health and Safety Group,
Confederation of British Industry (CBI).
Chris Ball Independent Consultant / Amicus-MSF.
Jan Berry Chairman,
Police Federation of England and Wales.
Petra Cook Chartered Management Institute
Karen Charlesworth
Valery Davey Former Member of Parliament for Bristol West.
Angela Ishmael The Andrea Adams Trust
Brian Kazer Chief Executive,
British Occupational Health Research Foundation.
Diana Lamplugh The Suzy Lamplugh Trust
Barbara Lindsay MBE Senior Policy Adviser,
Department of Trade and Industry.
Tom Mellish Trades Union Congress
David Palferman Health and Safety Executive
Scott O’Brien
Dianah Worman Chartered Institute of Personnel and Development.
Keith Handley
Imogen Haslam
Helge Hoel The University of Manchester
Sabir Giga
75
Appendix B: The Bullying Risk Assessment Tool (BRAT)
The following items relate to your experience within your organisation. Please rate each item
by circling the number that best corresponds to your experiences / thoughts over the last 6
months.
1
Strongly agree
2
Agree
3
Slightly agree
4
Slightly disagree
5
Disagree
6
Strongly disagree
1. New staff are made to feel welcome when starting employment in the
organisation
1 2 3 4 5 6
2. Conflict in my work unit is common 1 2 3 4 5 6
3. I am clear about what is expected from me 1 2 3 4 5 6
4. This organisation does not value equal opportunity for everyone 1 2 3 4 5 6
5. I have confidence in my line managers abilities 1 2 3 4 5 6
6. Staff shortages are common in my unit 1 2 3 4 5 6
7. I enjoy working in the teams that I am involved with 1 2 3 4 5 6
8. I am not clear about how to carry out my job 1 2 3 4 5 6
9. Vacant positions are filled without delay within my unit 1 2 3 4 5 6
10. My line manager tries to control every single aspect of what is going on at work 1 2 3 4 5 6
11. The grading / rank structure in this organisation is transparent 1 2 3 4 5 6
12. I don’t get on with some of my colleagues 1 2 3 4 5 6
13. I have received sufficient training to carry out my job 1 2 3 4 5 6
14. My unit often makes use of temporary staff 1 2 3 4 5 6
15. My line manager values constructive criticism 1 2 3 4 5 6
16. People in this organisation are not rewarded properly 1 2 3 4 5 6
17. I find my colleagues to be co-operative 1 2 3 4 5 6
18. I face conflicting demands in my job 1 2 3 4 5 6
19. Cover for absent staff is provided immediately within my unit 1 2 3 4 5 6
20. My line manager exploits his / her position of power 1 2 3 4 5 6
21. I feel my contribution to the organisation is recognised 1 2 3 4 5 6
22. Different professional groups don’t work well together within my unit 1 2 3 4 5 6
23. My job description is clearly defined 1 2 3 4 5 6
24. I feel that there isn’t enough time in the day to complete my work 1 2 3 4 5 6
25. My line manager consults me before decisions affecting me are made 1 2 3 4 5 6
26. The organisations’ resources are not distributed fairly 1 2 3 4 5 6
27. My line manager is sensitive to how I feel 1 2 3 4 5 6
28. Existing work pressures make it difficult to take time off work 1 2 3 4 5 6
29. Work is shared equally among the people I work with 1 2 3 4 5 6
76
Factor structure of BRAT factor structure
1. ‘Organisational fairness’: 1, 11, 21, 4, 16, 26
2. ‘Team conflict’: 2, 12, 7, 17, 22, 29
3. ‘Role conflict’: 8, 3, 18, 13, 23
4. ‘Workload’: 6, 14, 24, 28, 9, 19
5. ‘Leadership’: 5, 10, 15, 25, 27, 20
77
Appendix C: Validation of the BRAT (results of regression analysis)
Table A1: BRAT factors as independent predictors of NAQ-R
Beta t Sig.
BRAT – Org. fairness 0.129 3.310 P<0.001
BRAT – Team conflict 0.202 5.259 P=001
BRAT – Role conflict 0.155 4.216 P<0.001
BRAT – Workload (6) 0.107 3.174 P=0.002
BRAT - Leadership 0.207 5.637 P<0.001
Rsq =34.2%
Table A2: BRAT factors as independent predictors of NAQ-R - Work-related
harassment
Beta t Sig.
BRAT – Org. fairness 0.125 3.428 p=0.001
BRAT – Team conflict 0.115 3.311 p=0.001
BRAT – Role conflict 0.241 7.067 p<0.001
BRAT – Workload (6) 0.188 5.928 p<0.001
BRAT - Leadership 0.288 6.566 P<0.001
Rsq=41.2%
Table A3: BRAT factors as independent predictors of NAQ-R – Personal harassment
Beta t Sig.
BRAT – Org. fairness 0.116 2.755 p=0.006
BRAT – Team conflict 0.250 5.990 p<0.001
BRAT – Role conflict 0.055 1.383 NS
BRAT – Workload (6) 0.019 0.521 NS
BRAT - Leadership 0.162 4.019 p<001
Rsq=22.3%
Table A4: BRAT factors as independent predictors of self-labelled bullying
Beta t Sig.
BRAT – Org. fairness 0.129 3.119 p=0.002
BRAT – Team conflict 0.315 7.736 p<0.001
BRAT – Role conflict -0.099 -2.536 p=0.011
BRAT – Workload (6) 0.032 0.876 NS
BRAT - Leadership 0.192 4.848 p<001
Rsq=23.9%
Table A5: BRAT factors as independent predictors of General Health Questionnaire
(GHQ-12)
Beta t Sig.
BRAT – Org. fairness 0.066 1.584 NS
BRAT – Team conflict 0.107 2.565 p=0.11
BRAT – Role conflict 0.314 7.996 p<0.001
BRAT – Workload (6) 0.053 1.469 NS
BRAT - Leadership 0.100 2.488 p=0.013
Rsq-24.4%
78
Appendix D: Policy guidelines
NB: Bullet points in bold considered essential, other points considered desirable.
Statement of intent and commitment
• The right to work in an environment free of harassment, bullying &
intimidation (– working towards an environment free of
harassment/bullying).
• The seriousness of the problem, e.g. potential that disciplinary
actions may be taken, including dismissal. Can be unlawful/criminal
offence
• Applies to all employees, managers, workers, individuals
subcontracted or seconded to work for the organisation.
• The responsibility of all employees to comply with the policy
• Managers responsible for implementation of policies
• No recrimination/victimization
• Focus on a standard of conduct
Definitions & examples of behaviour and conduct in breach of policy
• Provide definition
• Examples of behaviours in breach of policy
• Highlights perception of recipient of exposure to unwanted
behaviour/bullying refers to behaviour which is unwanted by the
recipient
• Focus on deed as oppose to intent of behaviour acknowledging that
harassment and bullying can be unintentional
• (reference to a ‘reasonable’ person’s judgment of what constitutes
bullying
• Acknowledges/makes a distinction between one-off negative acts as
oppose to repeated negative behaviour
Principles for a safe complaint system: Reassurance of fairness, non-
recrimination and confidentiality
• No attempt of recrimination of targets (the person who files the
complaint) will be tolerated
• Confidentiality should be offered to the complainant (target) as far is
possible for the progression of the case. Unconditional
confidentiality cannot be offered as it may compromise the
employer’s general ‘duty of care’
• Principles of fairness and non-recrimination extends to alleged
perpetrators
• A message that complaints will be taken seriously
• Malicious complaints is considered a disciplinary offence
• Fairness to be assured by means of management training and
consistency in application of rules and regulations (reducing the impact of
subjectivity)
79
How to complain, seeking advice
• Line-managers should under normal circumstances be the first line
of contact
• Where to complain if the perpetrator happens to be the line-manager,
e.g. dedicated HR person/representative
• Emphasising the right to be accompanied when filing a complaint
• Outline the status and role of advisors and how the may be
contacted
o to provide advice on the rights of targets and the alternative
actions available to targets
o to provide practical help, e.g. assisting in drafting letters,
assisting targets to meetings etc.
• Professional support/counseling to be offered to target (complainant) as
well as alleged perpetrator throughout process
Reporting and complaints procedures
• Clear ad unambiguous procedures for resolving issues and
complaints
• Making a distinction between and informal and an informal complaint
rout
• Informal complaint (involving as few people as possible and with
interviews and discussions as far as possible held in an informal
atmosphere)
o Use of informal complaint does not prevent later use of the
formal complaint procedure
o Independence of investigators as far as possible
o Granting leave/suspension (with pay) during investigation if
necessary
• Formal complain
o Clarify how and where (to whom) complaint should be made
o Swift response, stating the given time period within which a
response will be made, e.g. a week
o Emphasise that the target should feel free to be accompanied
by a person of their choice in interviews,, e.g. a colleague or a
shop steward
o Potential sanctions listed
Information about how the policy is monitored (examples of potential approaches)
• Register of complaints/incidents (and their outcomes) to be retained
by HR
• Regular review of policy and monitoring system
• Collect information on negative behaviour as part of exit-interviews
• Include questions on the effectiveness of policy as part of staff surveys
80
Appendix E: Part A Tool: Bullying Risk Assessment Tool
Objective Hazard Identification: To be completed by the Human Resources
Department
1. Unit identification code: _______________
2. Total number of employees in unit_______
3. What is the average annual absenteeism rate in the unit*?
_______days per employee.
4. What is the average annual absenteeism rate in the organisation?
_______days per employee.
5. Is the unit experiencing specific cases of long-term absenteeism due to non-physical
illness?
Yes ? No ?
6. Are any employees in the unit often absent for shorter periods of time (i.e. 1 day at a time)
without explanation?
Yes ? No ?
7. What is the current annual employee turnover rate (percentage) in the:
(a) unit _______%
(b) organisation _______%
8. What was the annual employee turnover rate (percentage) in the previous year in the:
(a) unit _______%
(b) organisation _______%
9. How many complaints/grievances have been filed by employees from the unit in the last
year?
________
10. To what extent have the following affected the unit during the past 6 months:
Not at all
1
Some
2
Considerably
3
Extensively
4
Organisational / technological change 1 2 3 4
Financial budget cutbacks 1 2 3 4
Redundancies 1 2 3 4
Staff shortages / unfilled posts 1 2 3 4
Change of manager / supervisor 1 2 3 4
*Note: Unit refers to the level of analysis and will vary in each organisation (For example a ward in an
NHS Trust).
81
Appendix F: Demographic variations in the experience of bullying to bullying
Table A6: Gender differences in exposure to bullying
Male (%) Female (%) Sig.
Civil Service 9.2 15.3 p=.075
NHS 1 10.3 9.7 p=.499
NHS 2 12.5 17.1 p=.230
NHS 3 25.0 14.0 p=.152
Police 11.9 14.3 p=.313
Table A7: Age differences in exposure to bullying
Table A8: Contractual arrangement differences in exposure to bullying
Table A9: Hours of work differences in exposure to bullying
16-24
(%)
25-34
(%)
35-44
(%)
45-54
(%)
55-70
(%)
Sig.
Civil Service 13.3 8.6 14.9 15 12.1 p=.636
NHS 1 0 9.5 9.1 7.9 13.3 p=.923
NHS 2 0 22.7 13.3 13.2 25 p=.811
NHS 3 0 24.2 14.5 14.9 11.5 p=.936
Police 0 10.0 16.2 11.1 18.7 p=.789
Permanent
(%)
Fixed
period
(%)
Seconded
(%)
Other
(%)
Sig.
Civil Service 12.6 0 0 n/a p=1.000
NHS 1 11.0 0 0 0 p=1.000
NHS 2 15.9 25.0 n/a n/a p=.767
NHS 3 14.9 40.0 0 100 p=.002
Police 13.5 9.1 n/a 0 p=.997
Full-time (%) Part-time (%) Sig.
Civil Service 12.7 10.0 p=.633
NHS 1 10.9 6.8 p=.794
NHS 2 20.2 8.3 p=.382
NHS 3 20.6 7.4 p=.121
Police 12.3 18.2 p=.232
82
Table A10: Ethnic differences in exposure to bullying
Table A11: Organisational level of target
No
spvsr.
resp. (%)
Spvsr.
(%)
Middle
Mgmnt.
(%)
Snr.
Mgmnt.
(%)
Other
(%
Sig.
Civil Service 13.3 10.3 9.0 22.7 33.3 p=.243
NHS 1 14.7 8.9 6.2 0 5.3 p=.264
NHS 2 15.6 11.8 6.7 0 38.5 p=.433
NHS 3 13.5 15.4 24.2 33.3 12.5 p=.711
Police 11.9 17.4 17.2 0 14.3 p=.951
Table A12: Duration of bullying experience: breakdown within organisations
White (%) Non-white (%) Sig.
Civil Service 10.4 22.7 p=.010
NHS 1 9.8 10.5 p=.693
NHS 2 14.5 30.8 p=.317
NHS 3 12.8 34.4 p=.035
Police 12.2 33.3 p=.021
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Civil Service 32.3 22.6 22.6 22.6
NHS 1 5.3 15.8 21.1 57.9
NHS 2 9.5 23.8 38.1 28.6
NHS 3 23.5 23.5 29.4 23.5
Police 20.0 20.0 20.0 40.0
83
Table A13: Negative Acts: Breakdown within organisations
*(R)egularly or (O)ccasionally experiences
*
Someone withholding
information which affects
your performance (%)
Having your opinions and
views ignored (%)
Being exposed to
an unmanageable
workload (%)
O
45.9 37.7 34.7 Civil Service
R 8.4 7.3 11.9
O
47.0 37.0 36.2 NHS 1
R 11.9 8.7 14.1
O
48.6 45.3 42.1 NHS 2
R 7.9 8.7 10.7
O
44.8 46.2 30.3 NHS 3
R 10.5 8.1 17.5
O
49.0 38.0 29.3 Police
R 10.8 8.7 13.6
84
Appendix G: Pre- and post-intervention data comparisons
Table A14: Negative behaviour (no significant changes in terms of high
ranking negative behaviours)
*(R)egularly or (O)ccasionally experiences
Table A15: Duration of bullying experience
* Pre-
intervention
(%)
Post-
intervention
(%)
O 47.0 48.9 Someone withholding information which affects your
performance
R 10.0 9.4
O 40.4 40.6 Having your opinions and views ignored
R 8.2 7.3
O 33.8 34.7 Being exposed to an unmanageable workload
R 13.7 14.2
O 31.0 30.9 Being ordered to do work below your competence
R 9.4 11.0
O 29.2 29.9 Being given tasks with unreasonable or impossible
targets or deadline
R 8.7 10.3
O 24.5 25.1 Having key areas of responsibility removed or replaced
with more trivial or unpleasant tasks
R 6.9 6.6
O 22.5 22.5 Spreading gossip
R 4.8 5.2
O 22.4 23.2 Being shouted at or being the target of spontaneous
anger R 4.7 3.7
O 22.3 18.9 Being humiliated or ridiculed in connection with you
work
R 4.6 4.0
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Pre-intervention 20.0 21.5 25.9 32.6
Post-intervention 25.2 21.8 22.7 30.3
85
Table A16: Organisational breakdown of bullying duration
Within the
last 6
months (%)
Between 6 &
12 months
ago (%)
Between 1
and 2 years
ago (%)
More than
2 years
ago (%)
Civil Service (T0) 32.3 22.6 22.6 22.6
Civil Service (T1) 24.1 34.5 13.8 27.6
NHS 1 (T0) 5.3 15.8 21.1 57.9
NHS 1 (T1) 29.4 11.8 35.3 23.5
NHS 2 (T0) 9.5 23.8 38.1 28.6
NHS 2 (T1) 41.7 12.5 20.8 25.0
NHS 3 (T0) 23.5 23.5 29.4 23.5
NHS 3 (T1) 17.2 20.7 27.6 34.5
Police (T0) 20.0 20.0 20.0 40.0
Police (T1) 15.0 25.0 20.0 40.0
doc_888700757.pdf