You are on page 1of 25

The International Journal of Human Resource

Management

ISSN: 0958-5192 (Print) 1466-4399 (Online) Journal homepage: https://www.tandfonline.com/loi/rijh20

Measuring workplace bullying in organisations

Eleanna Galanaki & Nancy Papalexandris

To cite this article: Eleanna Galanaki & Nancy Papalexandris (2013) Measuring workplace
bullying in organisations, The International Journal of Human Resource Management, 24:11,
2107-2130, DOI: 10.1080/09585192.2012.725084

To link to this article: https://doi.org/10.1080/09585192.2012.725084

Published online: 17 Oct 2012.

Submit your article to this journal

Article views: 4727

View related articles

Citing articles: 14 View citing articles

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=rijh20
The International Journal of Human Resource Management, 2013
Vol. 24, No. 11, 2107–2130, http://dx.doi.org/10.1080/09585192.2012.725084

Measuring workplace bullying in organisations


Eleanna Galanaki* and Nancy Papalexandris

Department of Marketing and Communication, Athens University of Economics and Business,


Athens, Greece
This paper intends to explore alternative ways to measure workplace bullying and to
propose to HR professionals and academics a uniform and common way to assess the
phenomenon. Based on a survey on 840 junior and middle managers from diverse
sectors in Greece, we are trying to extract conclusions on the incidents and
characteristics of workplace bullying, in a country where empirical evidence on bullying
is very limited and where cultural dimensions differ from countries with extensive
research evidence on workplace bullying and wide application of preventive measures.
In Greece, workplace bullying is found to follow grossly similar patterns to those
reported in relevant studies in Europe. The results vary according to the measurement
methodology used. The latent class cluster analysis on the negative acts questionnaire
scale, as previously proposed by other authors, is found to reflect more accurately the
reality of workplace bullying occurrence, than other instruments (self-labelling or
operational methodologies, which are tested here). The conclusions of this study are
valuable to researchers and practitioners who wish to measure or compare the
occurrence of workplace bullying in their organisations, based on specific and
acceptable standards, around the globe.
Keywords: Greece; latent class cluster analysis; negative acts questionnaire;
workplace bullying

Introduction
The term ‘workplace bullying’ was introduced in the early 1990s, to represent a rather
common, yet overlooked phenomenon, with very serious repercussions on the well-being
of both employees and organisations. At the time, it was noticed that severe stress is
caused to employees by constant and repeated intimidating and demoting actions
occurring in their everyday life, which they had difficulty to confront. At the time, sexual
harassment was a much debated issue and the early researchers of workplace bullying
noticed the similarity between the two phenomena, especially in terms of frequency and
severity of impact on well-being (Einarsen 1999).
To give a definition, ‘Bullying takes place when one or more persons systematically
and over time feel that they have been subjected to negative treatment on the part of one or
more persons, in a situation which the person(s) exposed to the treatment have difficulty in
defending themselves against them. It is not bullying when two equally strong opponents
are in conflict with each other’ (Einarsen, Raknes and Matthı́esen 1994; Nielsen et al.
2009). As Einarsen put it, workplace bullying has two main features: repeated and
enduring aggressive behaviours (a) that are intended to be hostile and/or perceived as
hostile (b) by the recipient (Einarsen 1999).

*Corresponding author. Email: eleanag@aueb.gr

q 2013 Taylor & Francis


2108 E. Galanaki and N. Papalexandris

Under the above definition, the term bullying encompasses negative actions and
intimidation exerted by one or more persons (bullies). In the case that this repetitive,
negative treatment is exerted by a group of people, we specifically refer to it as ‘mobbing’
(Zukauskas and Vveinhardt 2011). In the work presented here, we will refer to bullying, as
originally broadly defined by Einarsen, so by ‘bullying’ we will also cover the act of
mobbing.
Workplace bullying is critical to organisations and employers that not simply want to
abide to legal restrictions and stipulations of the employment contract, but further wish to
fulfil the obligations dictated by the psychological contract. As multiple studies have
pinpointed, the respect and fulfilment of the psychological contract in any employment
relationship is crucial for multiple organisational behaviours and outcomes (Cohen 2011;
Nadin and Williams 2012; Tomprou, Nikolaou and Vakola 2012). By psychological
contract, we refer to: ‘a set of unwritten expectations concerning the relationship between
an employee and an employer. It is stated that the contract addresses factors that are not
defined in a written contract of employment such as levels of employee commitment,
productivity, quality of working life, job satisfaction, attitudes to flexible working, and the
provision and take-up of suitable training’ (Bloomsbury Business Library 2007, p. 6053).
In this sense, any workplace bullying phenomenon constitutes a breach of the
psychological contract, from the side of the employer, as he fails to guarantee the quality
of the working life and safeguard victims from workplace bullying. It can also be seen as a
breach of the contract from the part of the bullies against their employer, as they
undermine the quality of working life in the workplace.
The facets of the workplace bullying phenomenon are multiple and will be further
discussed later. Also the reported effects of the phenomenon on individuals (victims,
witnesses, even bullies) and organisations in which it occurs are multiple and very serious.
However, regardless of the vast amount of empirical research conducted in order to study
workplace bullying, almost 20 years after the first references to the phenomenon, we have
not yet concluded, how common this is and how many people are affected, even if most
authors agree that it is a problem that needs to be high in the agenda of most organisations.
As Nielsen et al. (2009) and Nielsen, Matthı́esen and Einarsen (2010) have recently
pinpointed, the vast majority of workplace bullying research reports much differentiated
results, even within Europe, possibly because of the difference in the measurement tools
used to assess bullying occurrence, but also due to country- and culture-specific
idiosyncrasies that are worth examining.
Given the undoubtedly intense negative effects of workplace bullying, the above-
mentioned difficulty in concluding to a unanimous extent on its occurrence and a uniform
way to measure it undermines the efforts of many professionals (line managers, HRM
professionals, consultants and organisational psychologists) who study and attempt to
diminish the phenomenon. By not being able to use sound, acceptable and comparable
evidence, as a way to benchmark and pinpoint weaknesses in the ‘battle’ against bullying,
these professionals find that their arsenal against such negative acts is undermined. For
example, which executive board would give orders to an HR manager to take measures
against bullying (measures that often cost money), if he is not able to prove that there
exists a problem with bullying at the organisational level, using comparisons, benchmarks
and standards?
It appears that, by not having concluded in a uniform way about tackling measurement,
people responsible for confronting the phenomenon at the organisational level lack the
appropriate credibility and their voice is often not being heard. The analyses conducted in
the current paper aim at adding up to the prevalent discussion on the recommended
The International Journal of Human Resource Management 2109

measurement methodology and conclude with some recommendations for HR


practitioners on how to measure workplace bullying and assess the severity of the
problem for their organisation, in order to take the necessary steps to confront it.

Literature review
Impact of bullying
The effects of workplace bullying that have been documented are manifold and affect both
individuals and organisations. Some of the reported effects on individuals include anxiety,
irritability, feelings of depression, paranoia, mood swings, feelings of helplessness, lowered
self-esteem, physical symptoms, social isolation and maladjustment, psychosomatic
illnesses, depressions, helplessness, anger, anxiety, despair, burnout and lowered
job satisfaction and well-being (Matthiesen and Einarsen 2004; Lutgen-Sandvik, Tracy
and Alberts 2007; Lutgen-Sandvik 2008). Many researchers have focused on the long-term
effects of bullying at work, notably post-traumatic stress disorder (Mikkelsen and Einarsen
2002; Matthiesen and Einarsen 2004). Some of the long-term effects of bullying focus on the
indirect impact on behaviours developed in the long term. One of the most characteristic
long-term outcomes is counter-aggression by the victim (Lee and Brotheridge 2006).
A somewhat different way to approach workplace bullying effects is from the
organisational standpoint. The effects of bullying are not limited to the victim(s) of bullying.
Much has been written about organisational or group effects, such as the quality of the
working environment (Einarsen et al. 1994) or the intention to leave (Djurkovic,
McCormack and Casimir 2008). On the other hand, organisations are seen as both
responsible for the incidence of bullying (Hoel and Cooper 2000; Salin 2003; Hauge,
Skogstad and Einarsen 2007; Coleman Gallagher, Harris and Valle 2008) and as the key to
reducing the frequency of the phenomenon or the severity of its impact (Rayner 1999;
Sheehan and Barker 1999; Coyne, Chong, Seigne and Randall 2003; Tehrani 2005; Hodson,
Roscigno and Lopez 2006; Health and Safety Authority 2007; Namie 2007; Bulutlar and Oz
2009; Salin 2009).

The need for measurement of the bullying phenomenon


The above-mentioned magnitude of the documented effects of workplace bullying
underlines the importance for organisations to monitor the occurrence of the phenomenon.
Organisations, especially in countries where employers are held accountable for the
bullying occurrence in the workplace, are forced to monitor negative acts and to proceed
responsibly in order to reduce them and foster a safe working environment. The first step
in dealing with any phenomenon is to elaborate on its magnitude through empirical
evidence. Academics, even in the traditional so-called ‘soft’ disciplines, are stressing the
importance of evidence in supporting any case (Beatty, Huselid and Schneier 2003; Pfeffer
and Sutton 2006), while HR metrics, in particular, have known a huge popularity in a
discipline that keeps attempting to prove its credibility (Murphy and Zandvakili 2000).
Regardless of the vast size of work on the effects and the preventive measures for
workplace bullying, and despite the many years since it first started to be discussed; the
biggest weakness in workplace bullying literature, research and preventive action is the
lack of a uniform way to tackle its measurement, which lowers the credibility of those
supporting the need for any preventive or corrective action.
Many recent papers deal with and focus on the issue of how we should measure
workplace bullying (see for example Notelaers, Einarsen, De Witte and Vermunt 2006;
2110 E. Galanaki and N. Papalexandris

Einarsen, Hoel and Notelaers 2009; Nielsen et al. 2010). At the same time, different
researchers adopt distinct measurement methodologies (Crawford 1997; Salin 2003;
Matthiesen and Einarsen 2004; Notelaers et al. 2006; Nielsen et al. 2009).
In a recent work, Nielsen et al. (2009) have performed a review on international studies
on workplace bullying and found that the percentage of bullying occurrence ranges from
1% to 55%, depending on the type of measurement used and the country in which the study
was realised. A recent meta-analysis of 102 published studies of workplace bullying has
also indicated a high difference in the reported occurrence of workplace bullying, by the
adopted measurement tool, depending on the country in which the study was conducted
(Nielsen et al. 2010).
In broad terms, there are two major approaches to the measurement of bullying:

The self-assessment method, whereby the respondents, usually based on a definition


provided to them, are asked to indicate whether they perceive themselves as victims of
bullying and to state how often they experience it. This approach has two main
characteristics. The one has to do with ease of use, as the categorisation of the
respondent as a bullying victim depends on his/her response to just one question. The
other characteristic of this methodology is the freedom it gives to respondents to decide
for themselves whether they are victims of bullying or not. This is compatible to the
notion of harassment, inherent to bullying, as harassment and negative treatment depend
on the perception of the victim. If they do not experience specific actions as bullying, it
should be up to them to decide whether they will name it ‘bullying’ or not. However, it
has repeatedly been found that this criterion reports lower rates of workplace bullying
than when the decision of categorising specific behaviours as bullying lies with the
researcher. This has been attributed to the face-saving inclination of many respondents,
‘recall problems, memory biases and distortions’, when a general question on bullying
is posed (Einarsen et al. 2009, p. 40), as well as to confidentiality issues, as ‘economic
dependency could effectively prevent people from being honest in their assessment,
especially when they are asked to assess the behaviour of their supervisors or people in
formal positions of power. (ibid., p. 26): it is one thing admitting that you have been
subjected to some specific negative acts, and a different story to admit being a victim of
workplace bullying.
The operational, behaviour- or criterion-based method (the most common instruments
used are the Negative Acts Questionnaire, NAQ) and the Quine inventory of
psychological terror. They provide a list of incidents which may arise within the working
environment, such as intimidation, constant critique of work and efforts, rumours, false
allegations and others. If the respondent reports a high frequency of occurrence of at
least one (or two, depending on the strictness of the measurement methodology) of the
incidents presented in the list, he/she is considered to be a victim of bullying. The
advantages of this method are that it leaves authority to the researcher to decide on the
categorisation, while it avoids reluctance and negative feelings that the respondent
might develop, which could create response bias. Therefore, this method appears more
objective than the self-reporting one. A problem with this method is that you might have
people reporting that they experience one or two negative acts very frequently, therefore
they are categorised as bullying victims, while they do not experience other negative
acts and they do not feel bullied (according to the self-report method). In our sample, for
example, we had people reporting that they are moved against their will very often, or
that they are constantly isolated. This, according to the operational criterion, would
The International Journal of Human Resource Management 2111

wrongfully result in being categorised as bullying victims, even if they do not experience
other negative acts frequently and they do not feel bullied.
A variation of the latter method was proposed recently, with the application of class
cluster analysis, in order to identify similar cases in the way they report negative acts
(types of negative acts and frequency of occurrence at the same time), therefore different
bullying experiences, instead of a ‘bullying/no bullying’ categorisation, based on the
frequency of one or two questionnaire items.

Importance of the setting: country and culture


In addition to the variety of instruments used to assess the occurrence of workplace
bullying, research on workplace bullying was also initiated and further developed in
specific countries, mostly Nordic and Anglo-Saxon, which led to a misrepresentation of
some parts of the world, in our effort to reach conclusions on the magnitude of the
phenomenon, globally (Nielsen et al. 2010). According to a recent thorough review of the
international empirical evidence on bullying, performed by Nielsen et al. (2009),
Scandinavian countries (Denmark, Finland, Norway, Sweden) have offered the largest
part of the global research on workplace bullying, while Anglo-Saxon countries (UK, USA
and Ireland) form the other pillar of workplace bullying research. German- and French-
speaking countries (Austria, Belgium, France) also have offered considerable empirical
evidence on workplace bullying, with the rest of the global empirical evidence on bullying
coming from somewhat geographically ‘disperse’ countries (Lithuania, Turkey and South
Africa). In addition to that, as previously mentioned, the results of these studies do not
agree with each other. We find a crucial aim emanating from these results: There is a need
for the adoption of a global, uniform research instrument that would help professionals
across the world, to assess the situation in their organisations. This necessity becomes even
more important for practitioners in multinationals, where standards and benchmarks need
to be applicable and meaningful across borders.

Aims and objectives


The current paper tries to investigate two broad sets of questions:
. How should we measure workplace bullying? Are some negative behaviours,
indicative to the existence of workplace bullying, more important to assess than
others?
. Which is the occurrence of workplace bullying in Greece? Do our findings agree with
findings from other European countries?

Research hypotheses
Under the light of the previously developed review on workplace bullying and the aims of
the current paper, we will attempt to test the following research hypotheses:
Hypothesis 1a: The measurement of workplace bullying through the operational
criterion will agree to the findings through the self-reporting technique
As mentioned above, many authors have pinpointed the differences between the
outcomes of the two methods (Zapf and Einarsen 2001; Nielsen et al. 2010). The main
advantage of the operational method, as already mentioned before, is that it leaves the
2112 E. Galanaki and N. Papalexandris

decision on whether someone is bullied, to the researcher (Notelaers et al. 2006).


However, this method has repeatedly been found, especially in Scandinavian countries, to
report higher levels of bullying than the self-reporting one (Mikkelsen and Einarsen 2001;
Salin 2001). This is not, however, the case in all the studies conducted in Europe, as in
some cases more respondents, on average, would self-report to be victims of bullying, than
found to be such from a list of behavioural experiences (like the NAQ or Quine
questionnaire) (Nielsen et al. 2010).
The self-reporting or subjective method, on the other hand, is easier to apply and gives
the freedom to the respondents to express their own views on whether they have
experienced bullying (Salin 2001). It also agrees to the notion of bullying itself: as in sexual
harassment, whether bullying has occurred or not should be in the eyes/feelings of the
victim. However, it is quite vague and perhaps difficult to discern for many respondents,
while it leaves much room to personal judgements that may be biased by most recent
experiences of the respondent. However, in general, the two methods are found to agree in
the majority of cases examined in previous studies.
Hypothesis 1b: The incidence of workplace bullying will be separated in two broad sets
of negative actions, i.e. one related to the person and one related to the
work that needs to be done
Many previous publications have repeatedly mentioned the existence of different sets
of bullying behaviours. One of the first who have mentioned this duality was Einarsen
(1999) who identified the existence of discrete negative behaviours which are related to the
work (work-related bullying) and others which are related to the victim (person-related
bullying). This was confirmed in posterior studies, where these two broad sets of negative
behaviours kept coming up, possibly not identically, but in a notionally consistent way
(Coyne et al. 2003; Beswick, Gore and Palferman 2006; Einarsen et al. 2009; Hoel, Glaso,
Hetland, Cooper and Einarsen 2010). Indicatively, Einarsen et al. (2009), and Notelaers
et al. (2006) consistently identify three factors within the NAQ questionnaire. One has to
do with negative actions that target the role and/or work of the employee, another on
negative actions targeting the person and one on the negative actions which aim to isolate
the employee.
Hypothesis 2: Greece is not expected to differ significantly in the occurrence of
workplace bullying from other countries in Europe, based on the
published findings of previous empirical studies
Greece has been traditionally considered as balancing between Europe and the near
East. The great Cretan writer Nikos Kazantzakis refers to ‘the two great currents which
constitute the double-born soul of Greece’ (Kazantzakis 1966, pp. 167 –168). By this
duality he meant the complex mixture of character constituents that have resulted from
Greece’s location between East and West and from the combination of classical with
modern elements in the Greek national character. However, in business life, Greece
appears to have adopted the way of doing things of the West. To this have contributed the
close relations of the country with Europe, indicative of which is its early participation to
the EU, as well as the very active presence in the economic being and doing of many
multinationals, most of which European or North-American. Empirical evidence on
workplace bullying in Greece is almost nonexistent, so there is no secondary sources that
could be used as an insight to what the Greek reality for workplace bullying is (Galanaki
and Papalexandris 2011). Therefore, we make the hypothesis that the findings on Greece
should not be considerably different to the ones from other European countries.
The International Journal of Human Resource Management 2113

Research methodology
Sample and data
A survey to 840 junior and middle managers was conducted in 2009. A group of research
assistants were instructed on the use of the questionnaire as a data collection technique and
were given contact details of line managers, whom they contacted with the aim of collecting
locally or at the respondent’s convenience, questionnaires on workplace bullying. The
questionnaires were given anonymously and the sole reservation was for the respondent to
hold a junior or middle management position, with a minimum of 2 years working
experience; 95.5% of the originally contacted managers completed the questionnaire,
while, thanks to the extensive instruction of the researchers, only 3 questionnaires were
rejected, leaving 840 total questionnaires. This more or less convenience sample is briefly
described in Table 1 that follows:
Table 1. Description of the 840-cases sample.
Profile of Age Average ¼ 39.24, SD ¼ 15.33
respondent
Gender Male 44%
Female 56%
Education Almost all respondents 99.8%
completed secondary education
Total years in education Average ¼ 15.4, SD ¼ 2.43
Experience Years of professional experience Average ¼ 15.82, SD ¼ 9.98
Years of supervising experience Average ¼ 4.93, SD ¼ 7.07
Profile of Size of company Number of employees Average ¼ 1372.4, SD ¼ 4266.8
company
Type of company Private 71.3%
Public or semi-public 28.7%
Greek 85.6%
Foreign 14.4%
Department Marketing/sales/supplies 17%
Production 7%
Finance/accounting/banking 26%
Services 44%
Human resources 4%

It can be noticed that it was attempted to represent in the sample people from different
backgrounds, as long as they fell within the middle and junior manager profile targeted at
the current research.

Measurement of bullying adopted by the authors


The authors of the present paper measured the occurrence of bullying with both
measurement approaches to workplace bullying, i.e. the self-assessment method and
the operational method (in this case, the NAQ-32 questionnaire). In both cases, the
respondents were asked to report whether he/she perceives that they are currently
the subject of bullying or negative acts. So, it should be noted that the respondents were
not asked to temporally define their bullying experience in terms of length of time or
specific time span. Instead, and in order to ease the response, they were asked whether they
are currently victims or receive negative acts, and further identified how often this occurs
(occasionally, at least once a month, at least once a week or daily).
. The self-assessment method, whereby the respondents, based on a definition
provided to them, were asked to indicate whether or not they perceive themselves as
2114 E. Galanaki and N. Papalexandris

victims of bullying and how often they experience it. The following definition of
bullying was given: ‘By workplace bullying we mean repeated and persistent
negative acts that are directed to one or more persons and create a negative working
environment. In bullying the targeted person has a difficulty in protecting and
defending him/her self. Therefore, bullying does not refer to conflicts between two
parties of equal strength’. Then the respondent was asked to evaluate whether he/she
considered oneself as a target of bullying according to the following five alternatives:
(1) no, (2) yes, occasionally, (3) yes, at least once a month, (4) yes, at least once a
week, (5) yes, every day. Respondents were considered to be victims of workplace
bullying, when they gave any answer of more frequent experience than ‘at least once
a week’ (equal or greater than 4). Therefore bullying under this measurement is
treated as a dichotomous variable of yes/no.
. The operational behaviour- or criterion-based method, specifically the NAQ, with
32-types of negative acts that may arise within the working environment, such as
intimidation, constant critique of work and efforts, rumours, false allegations and
others. The statements describing the different negative acts were followed by a scale
from (1) never, (2) yes, occasionally, (3) yes, at least once a month, (4) yes, at least
once a week, (5) yes, every day. In this measurement, a respondent is categorised as a
victim of bullying if he/she reports at least one negative act that occurs frequently (at
least once a week – equal or greater than 4).
. Class cluster operation on the NAQ. At the NAQ, we also applied a recently
proposed analytical methodology (Vermunt and Magidson 2002; Notelaers et al.
2006; Einarsen et al. 2009) named class cluster analysis which allows to group
together similar sets of responses. The advantage of this method is that it groups
people responding to a questionnaire (in this case the NAQ) based on the replies that
they gave on diverse questions, and which may be categorical, instead of continuous
(in this case frequency of negative acts incidents). In this sense, it groups people on
the proximity of their answers, taking into account that ‘never’ is a very different
answer from ‘occasionally’ and that the distance between them could not be
compared to the one between ‘occasionally’ and ‘once a month’. This analysis has
concluded that we may group the responses to three clusters. Although the inclusion
of a third cluster improved the two-cluster model characteristics, this is not the case
with the inclusion of a fourth cluster. The three-cluster model extracted is
characterised by very good indices of fit (L2 ¼ 22779.8, p , 0.01; BIC ¼ 20049.9;
dissimilarity index ¼ 0.970).

We named the three clusters to which we concluded, ‘no bullying’, ‘occasional


negative acts’ and ‘bullying’, based on the conditional probabilities in both tables. The
respondents of the ‘no bullying’ cluster had 86.33% probability of answering ‘never’ to
any NAQ question, which explains the name given to the cluster. The cluster ‘occasional
negative acts’ respondents had 63.2% probability of having answered ‘never’, but they
would report occasional negative acts at a 28.72% probability. Their probability of
mentioning more frequent negative acts was considerably lower (adding up to 8%). Finally
the respondents of the ‘bullying’ cluster would have only a 40% probability of answering
‘never’ to any question of NAQ, while they had a 20% probability of reporting a very
frequent negative act (at least once a week).
The characteristics of the three extracted clusters are depicted in Table 2, where the
average conditional probability of every type of answer is shown.
The International Journal of Human Resource Management 2115

Table 2. Mean conditional probabilities for the three class clusters, as a function of five answer
categories.
No bullying Occasional negative acts Bullying Total
(%) (%) (%) (%)
47.5 39.3 13.2 100
Never 53 41 6 100
Yes, occasionally 17 59 25 100
Yes, at least once a month 5 35 60 100
Yes, at least once a week 4 26 70 100
Yes, every day 4 18 77 100

The first row of Table 2 shows the percentage of cases that would fit into each cluster
and the following rows show the conditional probability of someone answering ‘never’ in
a question, to be categorised in cluster 1, cluster 2 or cluster 3.
Table 3, that follows, also shows the conditional probability of a respondent fitting to
each cluster, but in this case, what is actually shown is the conditional probability of a
respondent categorised as belonging to cluster 1, to have answered ‘never’, or ‘yes,
occasionally’, etc. to a question.

Table 3. Mean conditional probabilities for the five answer categories as a function of latent class
cluster type.

No bullying (%) Occasional negative acts (%) Bullying (%)


Never 86.33 63.26 40.36
Yes, occasionally 12.34 28.72 27.89
Yes, at least once a month 0.62 4.69 12.06
Yes, at least once a week 0.30 2.04 10.69
Yes, every day 0.44 1.36 9.60
100 100 100

Methods of analysis
Five different sets of analyses were applied on the data gathered, with IBM statistics 19
(previously SPSS):
. Descriptives, to make an inference on the occurrence of bullying in Greece and the
adoption or not of Hypothesis 2.
. Exploratory and confirmatory factor analysis, to analyse the structure of the NAQ
questionnaire and conclude on the acceptance or not of Hypothesis 1b.
. Correlations and cross-tabs, to establish if and how the NAQ measurements agree
with the subjective measurement of bullying (as assumed in Hypothesis 1a).
. Tests of difference of means for characteristics of ‘victims’, ‘occasional negative
acts’ and ‘non-victims’.

Findings
The NAQ scale: scores, reliability and composition
The NAQ scale has demonstrated an exceptional internal structural reliability, similar to
that reported in previous research projects conducted in other countries. Confirmatory
factor analysis indeed showed that all the 32 items are, as expected, grouped together very
2116 E. Galanaki and N. Papalexandris

well, while a very high Cronbach’s a (¼0.904) characterised the scale. This shows that all
items measure the same, or a similar concept, i.e. workplace bullying, while none
distances from the main underlying concept of negative acts (Table 4).
However, since in previous studies on bullying it has been proposed to identify
categories of negative acts, and in order to test our third Hypothesis, we decided to run an
exploratory factor analysis on the 32 items in order to see whether they group together
separately within the 32-item spectrum.
Our own results from the exploratory factor analysis were somewhat different from the
ones reported by Einarsen et al. (2009) and Notelaers et al. (2006), in which more factors
were extracted. However, the two factors mentioned in most previous studies, i.e. work-
related bullying and person-related bullying, appeared in our research as well. Further, the
application of factor analysis was shown to be a good choice, according to the Kaiser –
Mayer–Olin (KMO ¼ 0.912) and the Bartlett’s test of sphericity ( p ¼ 0.000), while the
eight factors extracted accounted for 58% of the total variance explained. We only had to
remove the variable ‘you are humiliated or ridiculed in connection with your work’,
because it loaded similarly to four factors, i.e. person-related bullying, criticism &
monitoring, intimidation, and racism. Therefore, we decided that, as the meaning of this
item is relevant to more types of negative actions, we would exclude it from the overall
factor analysis.
Specifically, we identified the following eight factors:
(1) Work-related bullying (eight items)
(2) Person-related bullying (four items)
(3) Criticism and monitoring (six items)
(4) Intimidation (six items)
(5) Racism (three items)
(6) Communication-related (two items)
(7) Location-related (two items)
(8) Sexual harassment (two items)
In order to further control for the existence of the two basic factors, as expected in
Hypothesis 1b, we run confirmatory factor analysis by randomly splitting the file and
running factor analysis in each of the two 420 cases datasets. Confirmatory factor analysis
gave identical outcomes to the ones from exploratory factor analysis. Specifically, the
eight factors mentioned above were identically extracted again in the two split files.

Occurrence of workplace bullying


In order to depict the occurrence of workplace bullying, based on each of the three
measurement strategies used in the current analysis, Table 5 shows the frequency of each
level of bullying occurrence according to our Greek sample. It seems obvious that the
operational methodology, according to which a person is seen as either a victim or not of
bullying, based on the frequency of occurrences of just one negative act, reports much
higher levels of workplace bullying than the other two methods of measurement. On the
other hand, the self-reporting method and the class cluster classification seem to conclude
to more similar workplace bullying levels.
Following that, we attempted to identify whether the people self-reporting as victims
of workplace bullying identify with the people that are found to be victims through the
operational measure of negative acts (at least one negative act reported to occur at least
once a week). We noticed that the operational criterion correlates strongly with the
Table 4. Rotated component matrix.

Work-related Person-related Criticism and Communication- Location- Sexual


bullying bullying Intimidation monitoring Racism related related harassment
You are ordered to do work clearly below 0.72 0.09 0.09 0.05 0.10 20.05 2 0.05 0.03
your level of competence
You are systematically required to carry 0.62 0.09 0.12 0.20 0.07 20.06 0.16 0.13
out tasks which clearly fall outside your job
description
Someone withholding information which 0.62 0.09 0.23 0.18 0.12 0.07 0.09 2 0.02
affects your performance
You are given tasks with unreasonable or 0.60 0.06 0.11 0.32 2 0.04 0.00 0.19 0.24
impossible targets or deadlines
Key areas of responsibility are removed or 0.59 0.29 0.14 0.21 0.01 0.19 0.07 2 0.14
replaced with trivial or meaningless tasks
You are ignored or excluded 0.57 0.41 0.01 0.10 0.02 0.22 2 0.08 2 0.32
Your opinions and views are ignored 0.52 0.10 0.22 0.30 0.07 0.02 2 0.08 2 0.06
You are exposed to an unmanageable 0.48 0.08 0.07 0.33 2 0.10 0.05 0.14 0.44
workload
You are subjected to false allegations 0.24 0.74 0.23 0.07 0.09 0.06 0.13 0.05
Rumours and gossip are spread about you 0.20 0.73 0.08 0.24 0.09 0.03 0.02 0.01
Insulting or offensive remarks about your 0.00 0.72 0.20 0.18 0.23 0.13 0.13 0.02
personal (e.g. habits and background) or
your private life
Somebody tries to sabotage your 0.18 0.57 0.27 0.13 2 0.02 0.04 0.30 0.08
performance
Hints or signals from others that you should 0.18 0.17 0.77 0.13 0.00 0.07 2 0.05 0.10
quit your job
Threats of making your life difficult (e.g. 0.25 0.16 0.70 0.10 0.05 0.13 0.05 2 0.01
The International Journal of Human Resource Management

overtime, unpopular tasks)


Intimidating behaviour, such as invasion of 0.00 0.24 0.61 0.02 2 0.02 0.33 0.23 0.08
personal space, shoving, blocking/barring
the way, etc.
(Continued)
2117
Table 4 – continued
2118

Work-related Person-related Criticism and Communication- Location- Sexual


bullying bullying Intimidation monitoring Racism related related harassment
Somebody causes you economic or 0.13 0.12 0.57 0.02 0.13 0.21 0.38 2 0.09
material damages
Pressure not to claim something which by 0.39 0.05 0.54 0.32 0.13 20.11 0.02 2 0.05
right you are entitled to (e.g. sick leave.
holiday entitlement, travel expenses)
You are the subject of excessive teasing 0.12 0.28 0.48 0.33 0.39 20.02 2 0.19 0.08
and sarcasm
You are repeatedly reminded of your errors 0.20 0.10 0.07 0.74 0.14 0.09 0.06 0.02
and mistakes
Persistent criticism of your work and effort 0.23 0.20 0.21 0.72 0.06 0.03 0.09 2 0.04
Your work is excessively monitored 0.23 0.05 0.03 0.69 2 0.05 0.09 0.13 0.09
You are shouted at or the target of rage or 0.21 0.25 0.14 0.49 0.07 0.03 2 0.03 2 0.09
anger
Insulting comments or behaviour with 0.14 0.13 0.06 2 0.04 0.74 0.03 0.13 0.05
reference to your religious or political
convictions
Physical abuse or threats of violence 2 0.06 20.12 0.12 0.14 0.70 0.17 0.25 0.03
Insulting remarks or behaviour with 0.12 0.28 2 0.02 0.09 0.64 0.09 2 0.10 2 0.02
reference to your native language, your
race or ethnicity
E. Galanaki and N. Papalexandris

You get insulting emails 0.10 0.11 0.14 0.08 0.08 0.83 0.01 2 0.08
You get other insulting written messages or 2 0.02 0.06 0.17 0.10 0.16 0.77 0.01 0.17
telephone calls
You are physically isolated 2 0.01 0.14 0.05 0.29 0.04 20.05 0.63 2 0.08
Being moved or transferred against your 0.25 0.22 0.10 2 0.08 0.25 0.06 0.58 0.07
will
You are sexually harassed or get unwanted 0.15 0.21 0.15 2 0.05 0.24 0.20 2 0.09 0.62
sexual attention
You are excluded from social events 0.35 0.33 0.17 0.02 0.15 0.14 0.03 2 0.48
Note: Bold values are the loadings of items on the corresponding factor, to which they were assigned.
The International Journal of Human Resource Management 2119

Table 5. Occurrence of workplace bullying according to each of the three measurement methods
used in the current study.
Bullying (%) Occasional negative acts (%) No bullying (%)
Self-reporting method 7.3 (at least 47 (less frequently 45.8
once a week) than once a week)
Class cluster method 13.2 39.3 47.5
Operational criterion 44.8 55.2
method (dichotomous)

self-reporting measure of bullying (Spearman correlations). However, this correlation,


being lower than 0.5, suggests that the two ways of measurement differ in their results.
This becomes more evident, when we apply cross-tab on the two alternative ways of
measurement, i.e. operational and self-reporting (see Table 6).
From Table 6 it becomes evident that the two criteria do not agree. Indicatively, of
those for whom the operational criterion has indicated that they are not victims of bullying,
39.5% report that they experience bullying at least occasionally, while of those self-
reporting that they have never been targets of bullying, 27% (or 12.4% of the total
respondents) appear, according to the operational criterion, to be categorised as victims of
repetitive negative actions or bullying. This obviously raises some questions regarding the
applicability of any of the two criteria and agrees with the discussion held by Notelaers
et al. (2006) and Einarsen et al. (2009).
In order to further examine the extent of agreement among the two methods, it was
decided to see at what percentage people being categorised in the cluster ‘never bullied’,
have answered ‘never’ or ‘yes occasionally’ to the self-reporting question on whether the
respondent has received bullying. Table 7 that follows presents the outcomes of this
analysis and shows how answers to each category (bullying, occasional bullying, no
bullying) agree between the two measurement methods.
From Table 7, it becomes apparent that the class cluster application on the operational
criterion increases the agreement between the two measurement methods. However, the
outcome of each method is quite different. This becomes more apparent, if we look at the
case of those answering that they receive bullying occasionally according to the definition
(less than once a week), but form 60.5% of those who were categorised as victims of
bullying according to the class structure methodology. Also, 31% of those saying that they
have never been bullied are categorised through the class cluster method to the occasional
bullying cluster. On the other hand, 22.7% of the people categorised in the never bullied
cluster would report that they experience bullying according to the definition,
occasionally.

Table 6. Cross-tab between self-reporting and operational measurement of bullying.

I have received bullying


Operational criterion Yes, Yes, at least Yes, at least Yes,
of bullying_1 NA Never occasionally once a month once a week every day
at least once a week (%) (%) (%) (%) (%) %
No bullying 60.5 35 3.2 1.1 0.2 100.0
Bullying 27.7 43.7 13.9 8.3 6.4 100.0
46 39 8 4 3 100
2120 E. Galanaki and N. Papalexandris

Table 7. Agreement between self-reporting and class cluster measurement methods (percent of
cases in each frequency category answer to the self-reporting method, by each of the three clusters).
Cluster modal
No bullying Occasional negative acts Bullying Total
(%) (%) (%) (%)
‘I have received bullying’
Never 76.1 31.0 7.7 45.8
Yes, occasionally 22.7 63.9 60.5 47.1
Yes, at least once a week 1.2 5.1 31.7 7.0
Total 100.0 100.0 100.0 100.0
39.6 47.2 13.2 100.0
Note: Bold values are the most astonishing groups of answers contradicting between the two types of measurement.

Finally, in order to see how the results of occurrence of bullying (cluster analysis) fit with
the categories of bullying identified through factor analysis, the operational criterion and the
self-report of bullying, we computed the correlations among the probability of belonging
to any of the three clusters, the self-reporting of being a victim of workplace bullying and
the eight factors of negative acts. The results are summarised in Table 8 that follows.
Interestingly, the subjective measurement of bullying seems to agree to a large extent
with the three clusters extracted. That is, for example, a person reporting being a victim of
bullying was less probable to be categorised in the ‘no bullying’ cluster (R ¼ 2 0.451,
p ¼ 0). On the other hand, some of the negative actions groups (factors) appeared to be
correlated with the bullying clusters and the self-reporting measurement, while others did
not. For example, work-related bullying, person-related bullying, and criticism and
monitoring agree perfectly with the three clusters and self-reporting measure through
significant and expected correlations. This is not the case with racism, sexual harassment,
and communication-related and location-related negative acts. It is only partly true in
intimidation, where there is a significant correlation with all but the probability to belong to
the ‘no bullying’ cluster.
This outcome, although perplexing at first, seems reasonable. Let us take the example of
location-related negative acts. Actually they do not exactly constitute bullying. They are
negative acts but scarcely will they be a repeated phenomenon, coupled with other negative
acts, which will accompany them. Actually, it should be noted that for the sake of the
current research, we used the older NAQ scale consisting of 32 items. Recently, a
shortened, revised version has been proposed, which does not include questions regarding
racism, sexual harassment and communication-related negative acts, as in the 32-item
NAQ used in the current survey (Einarsen et al. 2009), because such actions have been
shown to occur more scarcely; therefore, it is rare to have people reporting that the above
happen repeatedly, a major trait for a negative act to consist bullying according to the
bullying definition.
On the other hand, it should be noted that, as shown in Table 8, the experience of
bullying, as expressed in the self-labelling measurement, agreed more with ‘work-related
negative acts’ and ‘criticism and monitoring’, than ‘person-related bullying’.
Concurrently, ‘criticism and monitoring’ and ‘work-related bullying’ proved to be the
most common sets of negative acts, according to our sample. ‘Person-related bullying’ and
‘intimidation’, though significantly correlated with the self-nomination of a victim of
bullying, were found to occur less often. Table 9 demonstrates the average scores of each
of the eight negative acts factors adopted in the current study.
Table 8. Pearson correlation coefficients between the probability of belonging to each of the three clusters (identified through latent class cluster analysis),
operational criterion on bullying, self-reporting bullying and the eight groups of negative acts (identified through factor analysis).
Class cluster analysis
Operational criterion Self-labelling
on bullying method; ‘I have
No bullying Occasional negative acts Bullying (Spearman’s r) received bullying’
No bullying 1 2 0.754** 2 0.339** 20.567** 2 0.451**
Occasional negative acts 2 0.754** 1 2 0.362** 0.095** 0.113**
Bullying 2 0.339** 20.362** 1 0.625** 0.478**
Operational criterion on bullying (Spearman’s r) 2 0.567** 0.095** 0.625** 1 0.385**
‘I have received bullying’ 2 0.451** 0.113** 0.478** 0.385** 1
Work-related bullying 2 0.521** 0.197** 0.458** 0.597** 0.379**
Person-related bullying 2 0.195** 20.093** 0.410** 0.302** 0.177**
Intimidation 2 0.070 20.137** 0.294** 0.371** 0.162**
Criticism and monitoring 2 0.423** 0.165** 0.364** 0.610** 0.403**
Racism 2 0.053 20.074* 0.182** 0.167** 0.001
Communication-related 2 0.028 20.023 0.072* 0.140** 2 0.046
Location-related 2 0.084* 0.063 0.029 0.270** 0.081*
Sexual harassment 2 0.134** 0.091* 0.059 0.206** 0.086*
*
Significant at the 0.10 level.
The International Journal of Human Resource Management

**
Significant at the 0.05 level.
2121
2122 E. Galanaki and N. Papalexandris

Table 9. Average scores on each of the eight factors of negative acts.

Mean SD
Work-related bullying 1.7270 0.53406
Person-related bullying 1.4068 0.52033
Intimidation 1.1898 0.32077
Criticism and monitoring 1.8030 0.67030
Racism 1.0544 0.21867
Communication-related 1.0459 0.21843
Location-related 1.1967 0.41027
Sexual harassment 1.1647 0.33203
Valid N (listwise)

It appears then, that the most frequent negative acts are notably those that are related to
the type of work performed and to the control performed on the work provided, than the
person(s) involved or the interpersonal relations developed.

Setting of workplace bullying/negative acts


As repeatedly attempted in previous studies (Bilgel, Aytac and Bayram 2006; Hodson et al.
2006; Lutgen-Sandvik et al. 2007; Harvey, Treadway, Thompson and Duke 2008; Bryant,
Buttigieg and Hanley 2009; Bulutlar and Oz 2009; De Cuyper, Baillien and De Witte
2009; Salin 2009; Bjorkelo, Einarsen and Matthı́esen 2010; Hoel et al. 2010), we tried to
test whether individual or organisational characteristics are connected with the occurrence
of workplace bullying. We performed analysis of variance and Chi-square analyses,
depending on the type of variable in question, in order to test for the difference between
victims of bullying, occasional negative acts and no bullying. We did that for:
a. Individual characteristics of the respondent: age, gender, marital and family status,
working experience, hierarchical level, education level and behavioural traits, such as
conformity, work controls and confidence in work. Our analysis revealed that from all
the variables examined, only gender (x2 ¼ 6.639, df ¼ 2, p ¼ 0.036), the personal
attitudes of confidence in the work performed (F ¼ 14.130, df ¼ 2, p ¼ 0.000) and
work controls (F ¼ 9.168, df ¼ 2, p ¼ 0.000) of the respondent were significantly
different among victims and non-victims of bullying. Therefore, women were found to
receive bullying more than men, while bullying was less probable to occur to people
with confidence in the work performed and feelings of control over their work.
b. Organisational characteristics: private – public organisation, nationality of the firm,
department – activity and organisational culture of the department. Our analysis
revealed that from all the variables examined, only the power of organisational culture
in the eyes of the respondents was significantly different among victims, occasional
victims and non-victims of bullying.
This is a very interesting finding, as it agrees with previous studies, which hold that
there are no specific profiles of victims, while it is partly opposite to the finding of previous
studies that have underlined the critical importance of the organisational setting examined
(Einarsen et al. 1994; Hodson et al. 2006; Hoel and Beale 2006; Hauge et al. 2007; Hauge,
Skogstad and Einarsen 2009; Durniat 2010; Vickers 2010). Regarding gender, in
particular, the studies identifying one gender as more prone to receive bullying are scarce
(Lee 2002; Eriksen and Einarsen 2004; Ólafsson and Jóhannsdóttir 2004).
The International Journal of Human Resource Management 2123

Table 10. Percentage of respondents expressing that they have witnessed someone else being a
victim of workplace bullying.
Witnessing someone else being a victim of bullying %
Never 31.9
Yes, occasionally 42.2
Yes, at least once a month 9.5
Yes, at least once a week 10.2
Yes, daily 6.2

Additional findings, worth mentioning: witnesses of bullying and identity of bullies


In Tables 10 and 11, above, we present our findings regarding the experience of our
respondents in witnessing someone else receiving bullying, and their identification of the
person exerting bullying, accordingly. The percentage of respondents reporting that they
have witnessed others receiving bullying is comparable to the overall percentage of
bullying occurrence that we found here for Greece (16.4% are witnesses of bullying
occurring at least once a week).
Regarding the identity of the bully, it appears that, as in previous studies, the
immediate supervisor is most prone to bully his subordinates, while customers also very
commonly become bullies of the people who serve them. Colleagues from the same or
other department come next as possible bullies, whereas subordinates are less common to
become bullies, although this is something that actually happens to 2% of our respondents,
despite the difference in authority that should safeguard supervisors against being bullied
by their subordinates.

Discussion and hypotheses confirmation


The current paper focused on two major questions:
. How should we measure workplace bullying?
. Which is the occurrence of workplace bullying in Greece and are similarities found
with other European countries?
We will attempt to conclude for each of the two questions based on our findings.
Regarding our first question, our findings support the suggestion of Einarsen et al.
(2009) for the use of the NAQ questionnaire and the subsequent application of latent class
cluster analysis, in order to extract types of negative behaviour patterns. This method truly
appears to avoid the subjectivity of the self-reporting method and at the same time to take
into account the frequency dimension which is crucial in the definition of bullying. The
categories of replies were even more clear in the present research findings, as they were

Table 11. Frequency of the identity of each type of bully reported.

Immediate Colleague same Colleague other


Witnessing someone else supervisor department department Subordinate Customer
being a victim of bullying (%) (%) (%) (%) (%)
Never 57.8 70.3 73.7 85.2 67.1
Yes, occasionally 29.9 21.1 21.1 12.0 22.4
Yes, at least once a month 5.4 4.1 3.6 1.1 3.1
Yes, at least once a week 4.2 2.6 1.0 1.2 4.6
Yes, daily 2.8 1.8 0.6 0.6 2.9
2124 E. Galanaki and N. Papalexandris

restricted to only three categories, therefore the model extracted is more parsimonious
than the ones reported in previous studies (Notelaers et al. 2006; Einarsen et al. 2009).
With respect to Hypothesis 1a, according to which, ‘the measurement of workplace
bullying through the operational criterion will agree to the findings through the self-
reporting technique’, our analyses have indicated a very high correlation between the
bullying occurrence, as measured through the NAQ questionnaire, with the application of
the class cluster technique and the measurement through the self-reporting technique. This
positive correlation is not as high, between the traditional operational criterion (at least one
negative act per week) and the class cluster outcomes or the self-reporting technique. On
the other hand the two methods, even with the use of the class cluster analysis, seem to
give somewhat different results on some cases. Indicative of this is Table 5, according to
which, the incidence of bullying appears, in agreement to previous studies, to be higher if
measured through the class cluster technique (13.2%), as opposed to the self-reporting one
(7.3%), while the classification of several cases (bullying, occasional bullying, no
bullying) is different depending on the method used (see Tables 6 and 7).
The authors realise that the self-reporting technique has a critical advantage over the
behaviour or incidence techniques, such as the NAQ scale. This has to do with the freedom
given to the respondent to express his/her subjective perception of reality. This is very
important, since bullying, as sexual harassment, is ‘in the eyes of the victim’. However, it
appears that as in sexual harassment, respondents have difficulty to admit, even to
themselves, being victims of such behaviours, as they prefer a more face-saving strategy.
Therefore, we consider that we should accept Hypothesis 1a, stressing, though, that
under operational criterion, we recommend to researchers to use the NAQ scale, and then
apply to the data gathered class cluster analysis, instead of the more traditional method of
considering a respondent as a bullying victim, if he/she reports a more frequent than once a
week incidence of at least one negative act from the questionnaire.
The need for the application of the cluster analysis becomes even more obvious if one
thinks about the dichotomous nature of the operational criterion. This dichotomy between
bullied/not bullied, although useful for reasons of clear and straightforward presentation of
findings, proves inherently problematic in the sense that it misses to express the
‘occasional negative acts’ experience that is actually quite large according to our findings,
not to mention critical for the prevention of manifested bullying phenomena.
In relation to Hypothesis 1b, according to which ‘the incidence of workplace bullying
will be separated in two broad sets of negative actions, work-related and person-related
ones’, our findings, particularly through exploratory and confirmatory factor analysis,
indicated that both these two sets of negative actions, identified in numerous previous
studies, also came up in the present study. Therefore, we should accept Hypothesis 1b. As
a matter of fact, both work-related and person-related negative acts had a very high
correlation with self-reporting of workplace bullying incidence, which shows that they do
in fact constitute two faces of the bullying phenomenon.
However, it should be mentioned, in relation to this finding, that in our study more than
these two factors were extracted, but few factors were significantly correlated with the
levels of workplace bullying, as measured through both methods adopted in the current
study. The ones that were significantly correlated with the incidence of bullying were
work-related negative actions, person-related actions and criticism/monitoring. In fact,
people experiencing high criticism and monitoring or work-related negative actions were
more inclined than all others to self-report as victims of workplace bullying. It appears that
people are more prone to feel bullied due to work circumstances, expectations and
demands, as to the way their work should be offered, than by specific persons in their work
The International Journal of Human Resource Management 2125

environment. This is related to the common finding of previous empirical evidence, also
found in the current study, on the most common identity of the bully, being the immediate
supervisor (Høgh and Dofradottir 2001; Salin 2003; Harvey et al. 2007).
Regarding the second question, on the occurrence of workplace bullying in Greece, the
latent class cluster analysis has indicated that 13% of the respondents, or alternatively,
7.3%, according to the self-reporting technique, would fall within the ‘bullying victims’
category. This is similar to outcomes of previous studies in other European countries
(Nielsen et al. 2009, 2010), so Greece should not be considered as different in the
occurrence of bullying to what the international research has shown. However, this result
is accepted with caution, as the results of most previous studies are not consistent or
uniform. Our Greek sample has demonstrated a higher occurrence of bullying than some
previously reported results, conducted mainly in the Nordic countries, where there is both
a legal infrastructure protecting from bullying and a higher bullying conscience, due
mostly to the long academic interest in the topic.
Therefore, our findings, with reservations, support the acceptance of Hypothesis 2, at
large, according to which Greece was not expected to differ in the occurrence of workplace
bullying from other European countries, with the exception of some countries, notably the
northern ones. Indicative examples are Denmark and Finland, where it has been found at
2% and 1%, respectively, according to the self-reporting criterion, or Sweden, where it has
been found at 3.5% according to the operational criterion (Høgh and Dofradottir 2001; as
in Nielsen et al. 2009; Agervold 2007). In most European countries, findings are more
similar to the image depicted in the current Greek study, for example the French one,
where according to the self-reporting criterion 8.8% of the respondents are bullying
victims (Salin 2001), while in some countries the occurrence of workplace bullying is
more frequent than our Greek sample has indicated. Such are studies from the UK, Ireland
and Austria, where incidences of workplace bullying reach 10.6%, 23.1% and up to 26%,
respectively (O’Moore, Lynch and Niamh 2003; Hoel et al. 2010). It appears that Greece is
located in the middle of the reported workplace bullying occurrences, at the national level,
in Europe. Given that Greece is a country with non-existent legal stipulation on the matter,
or public debate on the issue, there is much room for actions to be taken on the matter,
mostly at the institutional and regulatory level.
Apart from the legal system and information on the phenomenon of bullying, societal
culture may also be considered as a proxy for the occurrence of workplace bullying. Societal
culture has been studied in several instances. Under business research, one of the most well-
known and recognised research projects on societal culture is the Global Leadership and
Organisational Behaviour Effectiveness (GLOBE) project. In relation to the current research,
we see ‘assertiveness’ and ‘power distance’ as the two cultural dimensions that most probably
could relate to the occurrence of workplace bullying. Greece was in fact found, through the
Globe study on 62 societies, to score quite high on both these dimensions (9th out of 62 in
assertiveness and 21st out of 62 countries in power distance) (Papalexandris 2007). In
addition to law and information, these cultural instances might also be a factor that encourages
bullying phenomena to occur, especially if the bully is the direct supervisor or someone higher
within the organisational hierarchy. Therefore, societal culture may explain the differences
found between our Greek sample and previous samples from Nordic countries.

Conclusions, limitations and suggestions for future research


The major outcome of the current study reflects to the country in which it was realised,
Greece. With 13% of our respondents being subjected to bullying (class cluster category),
2126 E. Galanaki and N. Papalexandris

the current research has shown that in Greece, as in the majority of Europe, workplace
bullying is an issue that needs to be looked at, both by organisations/employers and
institutions or government bodies. It should be noted that in Greece, workplace bullying is
not something for which the employee is protected by law. The employee can only plead
for the employer’s responsibility in offering him/her a safe and healthy working
environment. Therefore bullying is something employees can be protected from, legally,
only if their health is shown to be at risk through the negative or intimidating behaviour of
their predator.
The fact that the current study indicated a considerable occurrence of workplace
bullying is in fact even more alarming, if one thinks again about the current sample
structure. This study was addressed only to junior and middle level managers, therefore
people with certain power, esteemed by others, and with a quite high educational level.
One would expect that workplace bullying would be more common in professional classes
and lower hierarchical level employees who have more difficulty in defending themselves.
Having 13% of such a sample categorised as victims of workplace bullying is actually
alarming. Organisations expect these people to coach and manage others. If more than one
out of 10 has such experiences, which exactly is the management model that they will
adopt and what are the performance implications for the organisations employing them?
Management actions have a multiple effect on organisations, as they affect others.
Experience from studies in different countries has repeatedly shown that the
implications from this phenomenon are multiple and affect negatively many desirable
indicators for any organisation, such as coherence and collaboration, presence,
participation, focus to results, positive organisational culture, climate, health, low turnover,
etc. In addition to the above, we should bear in mind that the negative effects of bullying do
not keep to the internal of the organisation but reflect to its external environment from its
image as employer to its interaction with its main stakeholders, including the community as
a whole. Therefore, the extent of bullying occurrence reported here demands for close
monitoring and action to be taken in order to prevent such phenomena. There have been
numerous propositions for actions that organisations may take to prevent or reduce
workplace bullying, from training, counselling, grievance procedures, etc. (Bryant et al.
2009; Harlos 2010).
In the era of evidence-based management, though, the measurement tool by which an
organisation decides to take action on an emerging issue is almost as crucial as the
measures taken. How can the HR department propose to the top management to apply any
support or remedial programme for bullying, if it has not built a very strong case for the
magnitude of the phenomenon and its consequences in the organisation? The tool used is
of utmost importance. The second point one should keep from the current study has to do
with the way the phenomenon should be measured. Our main proposition is that the NAQ
is, as shown in numerous studies in the past, a most useful tool if coupled with the latent
class cluster analysis. It appears to give results very similar to ones of the self-report
method, while avoiding the perceived negative connotation or perceived risk in admitting
being a ‘workplace bullying’ victim. From the total scale, though, some questions do not
seem to be as critical as others; therefore particular attention should be attributed to work-
and person-related bullying, intimidation, and criticism/monitoring, which, according to
our findings appeared to be more significantly related to bullying in total, as well as most
commonly reported. A shortened version of the questionnaire, with only those four factors
(22 questions), is enough for an organisation to monitor bullying, unless other negative
behaviours are of specific interest to the given organisation. If class cluster analysis could
be used on the data gathered, the employer would be able to identify bullying victims and
The International Journal of Human Resource Management 2127

potential victims – occasional receivers of negative acts. These results should be used to
introduce a series of confidential and loose support policies for potential bullying victims.
An informal interview with an experienced professional from the HRM department would
most probably be the best way ahead in order to identify who is actually ‘bullying’ the
identified ‘victims’ and in what ways. The importance of the inside knowledge of the
organisation is critical in this respect as it is the experience of the HR professional in order
to avoid turning this process into a formal procedure that will miss its meaning. This initial
exploratory effort should be certainly followed by corrective action in case bullying is
actually confirmed.
A final conclusion from our analyses that should be stressed, even if not directly linked
to our research aims, has to do with the link of the organisational culture with the
occurrence of workplace bullying. Although based on a relatively small sample, our
results indicated that bullying was more probable to occur in organisations with stronger
cultural systems. This agrees perfectly with the findings of previous studies (Hodson et al.
2006; Hoel and Beale 2006; Hauge et al. 2007, 2009; Durniat 2010; Vickers 2010) and
should be taken in account by researchers and practitioners.
Of course, there are several limitations in the current study, the most obvious one being
the sample size (n ¼ 840), which is smaller than the ones in previously published studies,
as well as the structure of the sample, since all respondents matched the junior or middle
manager profile, therefore conclusions for manual or blue-collar workers in this specific
country are to be tackled with caution.
Another major limitation has to do with the time datedness of the data, as they were
collected in 2009. We anticipate that this time lapse affects the accuracy of our conclusions
especially for Greece, a country which, since late 2009, has entered in a fierce economic
crisis at the macro level that may have increased the experience of negative actions in the
workplace at the micro level.
Also, we should acknowledge that the research methodology does not allow us to
identify the length or the historical patterns of the bullying experience of respondents, as it
focused more on the current frequency of incidents. However, we believe that this is more
valuable to employers who wish to run a similar study in order to take preventive or
corrective action for bullying in their organisations.
Another limitation of this study, which however makes it interesting, is the fact that the
workplace bullying phenomenon is being studied for the first time in Greece, so there is no
possibility to do any comparison with primary or published information for the occurrence
of bullying in this specific country.
A suggestion for future research, based on the results of the current study, is to examine
workplace bullying, with the measurement tools suggested here, not only under the light of
specific organisations but also at the macro level. The recent and current Greek debt crisis,
for example has affected the way of doing business at large, while the subsequent structural
measures have dramatically altered the labour market at the macro level, directly affecting
employment practices and manifested stress in the workplace. It would be extremely
interesting to examine the effect of macro-elements, such as job security, unemployment
and income reduction, on the occurrence and experience of workplace bullying.
It would also be very useful to examine the connection of specific national culture
characteristics on the emergence of workplace bullying. Although bullying has been
related to the value systems of specific departments or organisations, the national culture
might also be a reason for the vast difference of reported bullying occurrence across
countries and this is a question worth exploring in future research. This demands, though
that a similar measurement instrument is adopted across countries.
2128 E. Galanaki and N. Papalexandris

References
Agervold, M. (2007), ‘Bullying at Work: A Discussion of Definitions and Prevalence, Based on an
Empirical Study,’ Scandinavian Journal of Psychology, 48, 161– 172.
Beatty, R.W., Huselid, M.A., and Schneier, C.E. (2003), ‘New HR Metrics: Scoring on the Business
Scorecard,’ Organizational Dynamics, 32, 107– 121.
Beswick, J., Gore, J., and Palferman, D. (2006), Bullying at Work: A Review of the Literature,
Derbyshire, UK: Health and Safety Laboratory.
Bilgel, N., Aytac, S., and Bayram, N. (2006), ‘Bullying in Turkish White-Collar Workers,’
Occupational Medicine, 56, 226– 231.
Bjorkelo, B., Einarsen, S., and Matthı́esen, S.B. (2010), ‘Predicting Proactive Behaviour at Work:
Exploring the Role of Personality as an Antecedent of Whistleblowing Behaviour,’ Journal of
Occupational and Organisational Psychology, 83, 371– 394.
Bloomsbury Business Library (2007), ‘Psychological Contract,’ in Bloomsbury Business Library –
Business & Management Dictionary, A&C Black Publishers Ltd, p. 6053.
Bryant, M., Buttigieg, D., and Hanley, G. (2009), ‘Poor Bullying Prevention and Employee Health:
Some Implications,’ International Journal of Workplace Health Management, 2, 48 – 62.
Bulutlar, F., and Oz, E.U. (2009), ‘The Effects of Ethical Climates on Bullying Behaviour in the
Workplace,’ Journal of Business Ethics, 86, 273– 295.
Cohen, A. (2011), ‘Values and Psychological Contracts in Their Relationship to Commitment in the
Workplace,’ Career Development International, 16, 646– 667.
Coleman Gallagher, V., Harris, J.K., and Valle, M. (2008), ‘Understanding the Use of Intimidation
as a Response to Job Tension: Career Implications for the Global Leader,’ Career Development
International, 13, 648– 666.
Coyne, I., Chong, P.S.-L., Seigne, E., and Randall, P. (2003), ‘Self and Peer Nominations of
Bullying: An Analysis of Incident Rates, Individual Differences, and Perceptions of the Working
Environment,’ European Journal of Work and Organizational Psychology, 12, 209– 228.
Crawford, N. (1997), ‘Bullying at Work: A Psychoanalytic Perspective,’ Journal of Community &
Applied Social Psychology, 7, 219– 225.
De Cuyper, N., Baillien, E., and De Witte, H. (2009), ‘Job Insecurity, Perceived Employability and
Targets’ and Perpetrators’ Experiences of Workplace Bullying,’ Work & Stress, 23, 206– 224.
Djurkovic, N., McCormack, D., and Casimir, G. (2008), ‘Workplace Bullying and Intention to
Leave: The Moderating Effect of Perceived Organisational Support,’ Human Resource
Management Journal, 18, 405– 422.
Durniat, K. (2010), ‘Organizational Culture’s Mechanism of Social Exclusion in the Process of
Workplace Bullying,’ in Proceedings of the European Conference on Management, Leadership
& Governance, 127– 134.
Einarsen, S. (1999), ‘The Nature and Causes of Bullying at Work,’ International Journal of
Manpower, 20, 16 – 27.
Einarsen, S., Raknes, B.I., and Matthı́esen, S.B. (1994), ‘Bullying and Harassment at Work and
Their Relationships to Work Environment Quality: An Exploratory Study,’ European Work and
Organizational Psychologist, 4, 381– 401.
Einarsen, S., Hoel, H., and Notelaers, G. (2009), ‘Measuring Exposure to Bullying and Harassment
at Work: Validity, Factor Structure and Psychometric Properties of the Negative Acts
Questionnaire-Revised,’ Work & Stress, 23, 24 – 44.
Eriksen, W., and Einarsen, S. (2004), ‘Gender Minority as a Risk Factor of Exposure to Bullying at
Work: The Case of Male Assistant Nurses,’ European Journal of Work and Organizational
Psychology, 13, 473– 492.
Galanaki, E., and Papalexandris, N. (2011), ‘Exploring Workplace Bullying in Greece: Frequency of
Occurrence and Handling of Measurement Issues, Under the Light of Previous Research
Findings at the International Level,’ in European Academy of Management 2011 Conference:
Management Culture in the 21st Century, Tallinn, Estonia, June 1 – 4.
Harlos, K. (2010), ‘If You Build a Remedial Voice Mechanism, Will They Come? Determinants of
Voicing Interpersonal Mistreatment at Work,’ Human Relations, 63, 311– 329.
Harvey, M., Buckley, M.R., Heames, T.J., Zinko, R., Brouer, L.R., and Ferris, R.G. (2007), ‘A Bully
as an Archetypal Destructive Leader,’ Journal of Leadership & Organizational Studies, 14,
117– 129.
Harvey, M., Treadway, D., Thompson, H.J., and Duke, A. (2008), ‘Bullying in the 21st Century
Global Organization: An Ethical Perspective,’ Journal of Business Ethics, 85, 27 – 40.
The International Journal of Human Resource Management 2129

Hauge, L.J., Skogstad, A., and Einarsen, S. (2007), ‘Relationships Between Stressful Work
Environments and Bullying: Results of a Large Representative Study,’ Work & Stress, 21,
220– 242.
Hauge, L.J., Skogstad, A., and Einarsen, S.L. (2009), ‘Individual and Situational Predictors of
Workplace Bullying: Why do Perpetrators Engage in the Bullying of Others?’ Work & Stress,
23, 349– 358.
Health and Safety Authority (2007), Code of Practice for Employers and Employees on the
Prevention and Resolution of Workplace Bullying, Dublin: Health and Safety Authority.
Hodson, R., Roscigno, J.V., and Lopez, H.S. (2006), ‘Chaos and the Abuse of Power: Workplace
Bullying in Organizational and Interactional Context,’ Work and Occupations, 33, 382– 416.
Hoel, H., and Beale, D. (2006), ‘Workplace Bullying, Psychological Perspectives and Industrial
Relations: Towards a Contextualized and Interdisciplinary Approach,’ British Journal of
Industrial Relations, 44, 239– 262.
Hoel, H., and Cooper, C.L. (2000), Destructive Conflict and Bullying at Work, Manchester: Launch
of the Civil Service Race Equality Network.
Hoel, H., Glaso, L., Hetland, J., Cooper, C.L., and Einarsen, S. (2010), ‘Leadership Styles as
Predictors of Self-Reported and Observed Workplace Bullying,’ British Journal of Manage-
ment, 21, 453– 468.
Høgh, A., and Dofradottir, A. (2001), ‘Coping With Bullying in the Workplace,’ European Journal
of Work and Organizational Psychology, 10, 485– 495.
Kazantzakis, N. (1966), Travels in Greece, Oxford: Bruno Cassirer.
Lee, D. (2002), ‘Gendered Workplace Bullying in the Restructured UK Civil Service,’ Personnel
Review, 31, 205– 227.
Lee, R.T., and Brotheridge, C.M. (2006), ‘When Prey Turns Predatory: Workplace Bullying as a
Predictor of Counteraggression/Bullying, Coping, and Well-Being,’ European Journal of Work
and Organizational Psychology, 15, 352– 377.
Lutgen-Sandvik, P. (2008), ‘Intensive Remedial Identity Work: Responses to Workplace Bullying
Trauma and Stigmatization,’ Organization, 15, 97– 119.
Lutgen-Sandvik, P., Tracy, J.S., and Alberts, K.J. (2007), ‘Burned by Bullying in the American
Workplace: Prevalence, Perception, Degree and Impact,’ Journal of Management Studies, 44,
837– 862.
Matthiesen, S.B., and Einarsen, S. (2004), ‘Psychiatric Distress and Symptoms of PTSD Among
Victims of Bullying at Work, British Journal of Guidance & Counselling,’ British Journal of
Guidance & Counselling, 32, 335–356.
Mikkelsen, E.G., and Einarsen, S.L. (2001), ‘Bullying in Danish Work-Life: Prevalence and Health
Correlates,’ European Journal of Work and Organizational Psychology, 10, 393– 413.
Mikkelsen, E.G., and Einarsen, S. (2002), ‘Basic Assumptions and Symptoms of Post-Traumatic
Stress Among Victims of Bullying at Work,’ European Journal of Work and Organizational
Psychology, 11, 87 – 111.
Murphy, T.E., and Zandvakili, S. (2000), ‘Data- and Metrics-Driven Approach to Human Resource
Practices: Using Customers Employees, and Financial Metrics,’ Human Resource Management,
39, 93.
Nadin, S.J., and Williams, C.C. (2012), ‘Psychological Contract Violation Beyond an Employees’
Perspective: The Perspective of Employers,’ Employee Relations, 34, 110– 125.
Namie, G. (2007), ‘The Challenge of Workplace Bullying,’ Employment Relations Today, 34, 2,
43 – 51.
Nielsen, M.B., Matthı́esen, S.B., and Einarsen, S. (2010), ‘The Impact of Methodological
Moderators on Prevalence Rates of Workplace Bullying. A Meta-Analysis,’ Journal of
Occupational and Organizational Psychology, 83, 955– 979.
Nielsen, M.B., Skogstad, A., Matthiesen, S.B., Glasø, L., Aasland, M.S., Notelaers, G., and Einarsen,
S. (2009), ‘Prevalence of Workplace Bullying in Norway: Comparisons Across Time and
Estimation Methods,’ European Journal of Work and Organizational Psychology, 18, 81 –101.
Notelaers, G., Einarsen, S., De Witte, H., and Vermunt, J.K. (2006), ‘Measuring Exposure to
Bullying at Work: The Validity and Advantages of the Latent Class Cluster Approach,’ Work &
Stress, 20, 289– 302.
O’Moore, M., Lynch, J., and Niamh, N.D. (2003), ‘The Rates and Relative Risks of Workplace
Bullying in Ireland, a Country of High Economic Growth,’ International Journal of
Management and Decision Making, 4, 82 – 95.
2130 E. Galanaki and N. Papalexandris

Ólafsson, R.F., and Jóhannsdóttir, H.L. (2004), ‘Coping With Bullying in the Workplace: The Effect
of Gender, Age and Type of Bullying,’ British Journal of Guidance & Counselling, 32,
319– 333.
Papalexandris, N. (2007), ‘Greece, From Ancient Myths to Modern Realities, Ch. 21,’ in Culture and
Leadership Across the World: The GLOBE Book of In-Depth Studies of 25 Societies, eds. J.S.
Chhokar, F.V. Brodbeck, and R.J. House, Mahaw, NJ: Lawrence Erlbaum Associates,
pp. 767– 802.
Pfeffer, J., and Sutton, I.R. (2006), ‘Evidence-Based Management,’ Harvard Business Review,
January, 62– 74.
Rayner, C. (1999), ‘From Research to Implementation: Finding Leverage for Prevention,’
International Journal of Manpower, 20, 28 – 38.
Salin, D. (2001), ‘Prevalence and Forms of Bullying Among Business Professionals: A Comparison
of Two Different Strategies for Measuring Bullying,’ European Journal of Work and
Organizational Psychology, 10, 425– 441.
Salin, D. (2003), ‘Ways of Explaining Workplace Bullying: A Review of Enabling, Motivating and
Precipitating Structures and Processes in the Work Environment,’ Human Relations, 56,
1213– 1232.
Salin, D. (2009), ‘Organisational Responses to Workplace Harassment: An Exploratory Study,’
Personnel Review, 38, 26– 44.
Sheehan, M., and Barker, M. (1999), ‘Applying Strategies for Dealing With Workplace Bullying,’
International Journal of Manpower, 20, 50 – 56.
Tehrani, N. (2005), Bullying at Work: Beyond Policies to a Culture of Respect, London, UK:
Chartered Institute of Personnel and Development.
Tomprou, M., Nikolaou, I., and Vakola, M. (2012), ‘Experiencing Organizational Change in Greece:
The Framework of Psychological Contract,’ International Journal of Human Resource
Management, 23, 385– 405.
Vermunt, J.K., and Magidson, J. (2002), ‘Latent Class Cluster Analysis,’ in Applied Latent Class
Analysis, eds. J.A. Hagenaars and A.L. McCutcheon, Cambridge: Cambridge University Press,
pp. 89 – 106.
Vickers, M.H. (2010), ‘Introduction – Bullying, Mobbing, and Violence in Public Service
Workplaces,’ Administrative Theory & Praxis (M.E. Sharpe), 32, 7 – 24.
Zapf, D., and Einarsen, S. (2001), ‘Bullying in the Workplace: Recent Trends in Research and
Practice – An Introduction,’ European Journal of Work and Organizational Psychology, 10,
369– 373.
Zukauskas, P., and Vveinhardt, J. (2011), ‘Mobbing Diagnosis Instrument: Stages of Construction,
Structure and Connectedness of Criteria,’ Journal of Business Economics & Management, 12,
400– 416.

You might also like