i 
 

 

 

STUDENT NAME: Nomsa Malatsi 

STUDENT NUMBER: 755577  
 

Bullying behaviour and psychosomatic health among township learners: Cross-

sectional survey. 
 

SUPERVISOR: Dr Sibusiso Mkwananzi 

 

Research report submitted to the University of the Witwatersrand, in the Human and 

Community Development Department (Faculty of Humanities), in partial completion of 

the requirements for the degree of Master in Educational Psychology, Johannesburg, 

March 2020. 

 

 

  

 

 



  

ii 
 

Declaration  

 

I hereby declare that this research thesis is my own work and has never been submitted or 

published before for any other degree qualification or examination at any academic institution. 

 

This research is being submitted to the University of Witwatersrand, Johannesburg for the 

degree Master of Educational Psychology. 

 

 

 

 

-------------------------------------                                         ------------------------------------- 

Nomsa Malatsi                                                                   Date 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

iii 
 

ACKNOWLEDGEMENTS  

 

I would like to extend my sincere gratitude to my supervisor Dr Sibusiso Mkhwananzi for her 

guidance, support, encouragement, and patience; thank you very much. To my first supervisor 

Dr Joseph Seabi, no words can describe how thankful I am for your support and patience. 

Thank you for assisting me during the early stages of my research study. 

 

Secondly, I would like to thank the Gauteng Department of Education for giving me permission 

to conduct this study in the schools in the province, and also the principals of the three Orlando 

West high schools who allowed me onto their school premises and granted me access to work 

with their learners. I truly appreciate your support and willingness to assist with the study. 

Additionally, I would like to thank the learners from these three schools for taking time to 

participate in my study; if it were not for you, this study would not have been possible. Ke ya 

leboha haholo.  

 

Ms Khoza and Professor Diale, thank you for mentoring me from the beginning of my 

Educational Psychology journey till this day. I am truly grateful and blessed to have met you.  

 

I would also like to thank Mr Vincent Siwawa, Professor Brown, Ms Portia Makhoba, and Ms 

Gillian Godsell for assisting me during the early stages of my research study and always 

pushing and encouraging me when I felt like giving up. 

 

To my older brother Wandile, thank you very much for your sacrifices and encouraging words, 

and always believing in me, and also for making it possible that I get to the schools on time to 

conduct my research. 

 

To my parents Patrick and Maria Malatsi, little brother Zamokuhle Malatsi and my partner 

Gomolemo Rapeane, thank you so much for always being there to offer me your ears and time 

when I needed a shoulder to cry on. 

 

Lastly, I would like extend a word of deep gratitude to my friend Thato Maroga for assisting 

me with data collection from the schools. I truly appreciate your help and support. 

MAY GOD BLESS YOU ALL ABUNDANTLY!!! 

 



  

iv 
 

 

ABSTRACT  

 

The study aimed to establish the association between bullying and psychosomatic health. The 

target population for this study consisted learners from three Orlando West high schools in the 

township of Soweto, Johannesburg. Approximately 300 learners were sampled, comprising 

girls and boys from grades 9–11. A convenient sampling technique was used to select 

participants for this research. The study analysed data from respondents using univariate, 

bivariate and multivariate analysis, specifically descriptive statistics as well as logistic 

regression. Regarding bullying, the study concluded that it happened more among boys than 

girls. Psychosomatic health challenges, however, occurred more among girls than boys and 

there was an association between bullying and psychosomatic health. The study has shown that 

bullying can affect the psychosomatic health of young people beyond psychological and 

academic performance effects as seen in previous studies. Going forward, it would be important 

for schools to prevent and decrease the levels of bullying in order to ensure increased levels of 

psychosomatic health among school children in South Africa.  

 

 

Keywords: Bullying, psychosomatic health, South Africa, gender differences 

 

 

                            

 

 

 

 

 

 

 

 

 

 

 



  

v 
 

LIST OF FIGURES 

Figure 4.1: Gender distribution across sample                                                                        27 

Figure 4.2: Education level                                                                                                     27 

Figure4.3: Participants’ home languages                                                                                28 

Figure 4.4: Levels of bullying                                                                                                 28 

Figure 4.5: The prevalence of psychosomatic health conditions in township schools            29 

Figure 4.6: Difficulty concentrating                                                                                        29 

Figure 4.7: Difficulty sleeping         30 

Figure 4.8: Headache                                                                                                               30 

Figure 4.9: Stomach ache                                                                                                        31 

Figure 4.10: Feeling tense                                                                                    31 

Figure 4.11: Loss of appetite                                                                                                   32 

Figure 4.12: Feeling sad                                                                                                          32 

Figure 4.13: Feeling dizzy                                                                                                       33 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

vi 
 

LIST OF TABLES 

Table 4.1: Levels of types of bullying by gender                                                                    34 

Table 4.2: Levels of psychosomatic health conditions                                                            35 

Table 4.3: Regression results                                                                                                   36 

Table 4.4: Relationship between being bullied and psychosomatic condition   37 

Table 4.5: Association between being a bully-victim and psychosomatic condition   38 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

vii 
 

TABLE OF CONTENTS 

 

CONTENT          i 

DECLARATION         ii  

ACKNOWLEDGEMENT        iii  

ABSTRACT          iv  

LIST OF FIGURES                               v 

LIST OF TABLES                                                                vi  

TABLE OF CONTENTS        vii 

 

Chapter 1: Introduction………………………………………………………….  1 

1.1 Introduction……………………………………………………………………. 1 

1.2 Background……………………………………………………………………. 1 

1.3 Problem statement……………………………………………………………..  3 

1.4 Rationale………………………………………………………………………. 5 

1.5 Aims of the study……………………………………………………………...  6 

1.6 Research questions…………………………………………………………….  6 

1.7 Research process……………………………………………………………… 6 

1.7 Outline of the chapters………………………………………………………… 7 

 

Chapter 2: Literature review……………………………………………………. 8 

2.1 Introduction……………………………………………………………………. 8 

2.2 Bullies…………………………………………………………………………. 9 

2.3 Victims……………………………………………………………………….....10 

2.4 Gender difference and bullying……………………………………………….. 11 

2.5 Prevalence of bullying/victimisation………………………………………….. 12 

2.6 Traditional bullying and cyberbullying……………………………………….. 12 

2.7 Consequences of bullying…………………………………………………….   13 

2.8 Health problems……………………………………………………………….  14 

2.9 The South African legislation on bullying…………………………………….. 14 

2.10 The constructs between Bullying and Psychosomatic Health.………….......... 17 

2.11. Theoretical framework…………………………………………………….   17 

2.11.1 Social Dominance Theory.………………………………………………… 17 

 

 

 



  

viii 
 

Chapter 3: Methods……………………………………………………………… 19 

3.1 Introduction……………………………………………………………………. 19 

3.2 Theoretical framework…………………………………………..………….… 19 

3.3 Research paradigm.………………………………………………………….... 19 

3.4 Research design…………………………………………………..…………… 20 

3.5 Participants…………………………………………………....……………....  20 

3.6 Instruments.…………………….……………………………………………..  21 

3.7 Procedure………… …………………………………………………………..  23 

3.8 Data analysis..…………………………………………………..…………….. 24 

3.9 Ethical considerations………………………………………………………….24 

 

Chapter 4: Results………………………………………………………………..  26 

 4.1 Introduction…………………………………………………………………..   26 

 4.2 Descriptive outcome.………………………………………………………..    26 

 4.3 Inferential outcome………………………………..…………………………   33 

 

Chapter 5: Discussion……………………………………………………………  39 

5.1 Levels of psychosomatic conditions…………………………………………..  39 

5.2 Levels and gender differentials of bullying…………………………………...  40 

5.3 Association between bullying and psychosomatic conditions………………...  42 

5.4 Limitations and future direction……………………………………………….  44 

5.5 Conclusion…………………………………………………………….……… . 45 

 

6.0 Reference list……………………………………………………….………….  48 

7.0 Appendices…………………………………………………………………….  57



  

1 
 

CHAPTER 1: INTRODUCTION 

 

“Schools are war-zones” (Harber, 2001:262) 

 

1.1 Introduction 

 

Violence is the deliberate use of physical influence or force against oneself or another person 

or against a group or community that has a high likelihood of resulting in death, psychological 

harm, maldevelopment, injury, or deprivation (Burton & Leoschut, 2013). In South Africa, 

there is a high death rate of 6.8 per 100 000 of children aged 10–19 years, which is more than 

four times the rate for adolescent females with some regional variation (Mathews, et al., 2019). 

The high death rate is caused by public violence and domestic injuries. With a high rate of 

violence in the home, it may not be surprising that school violence is among the most pressing 

concerns that the South African education system faces today and is thus associated with the 

way young people experience school (Crawage, 2008; Maree, 2005). According to Burton and 

Leoschut (2013), school violence is defined as violence that occurs within the physical border 

of the school environment.  

 

1.2 Background 

 

Globally, research focusing specifically on school bullying has undergone four phases (Smith, 

2012). The first phase dated from the 1970s to 1988. In 1978, the Swedish-Norwegian 

psychologist Dan Olweus published Aggression in schools: Bullies and whipping boys. In this 

book, bullying was defined in terms of mobbing, physical, and verbal behaviours, although 

Olweus (1978) explicitly rejected the ‘mobbing’ label (which implies group bullying), since 

much bullying appeared to be by one person (Søndergaard, 2020). The second phase was from 

around 1989 to the mid-1990s when more books as well as journal articles started to appear, 

and surveys in other countries beyond Scandinavia were beginning to be carried out. This was 

further helped by some intervention campaigns to fight against school bullying in Scandinavian 

countries. The third phase led to an established international research programme from the mid-

1990s to 2004, in which “traditional bullying was recognised as an important international 

research program” (Smith, 2013:71). Surveys and interventions took place in many countries 

(Smith, Pepler, & Rigby, 2004) and researchers in the United States developed substantial 

research on victimisation and bullying during this period (Espelage & Swearer, 2004), while 



  

2 
 

important work was also being undertaken in Australia and New Zealand (Rigby, 2002). The 

fourth phase was specifically focused on cyberbullying from 2004 to date. Over the past 

decade, cyberbullying has become a significant aspect to be taken account of (Smith, 2012). 

Rivers and Noret (2010:646) argue:  

This started off as text message and email bullying, which increased through the mid-

2000s; but since then the development of smart phones, and greatly increased use social 

networking sites, have offered many new tools for those wishing to hurt others. Broadly 

speaking, cyberbullying now takes up an appreciable fraction of total bullying in young 

people (around one-third). 

 

School violence results in a violation of the basic rights of children as it affects learners’ 

academic performance, increases isolation, and may also place both their mental and physical 

health at risk (Burton, 2008). School violence instils fear and negatively impacts children who 

are not directly experiencing violence but may have seen their friends or peers affected (Burton 

& Leoschut, 2013; Taole, 2016). According to Buston (cited in Ncontsa & Shumba, 2013), in 

South African primary and secondary schools a total of 1.8 million pupils (which is 15.3% of 

learners) had experienced violence in one form or another; 12.8% had been threatened with 

violence; 5.8% had been assaulted; 4.6% had been robbed of their belongings; and 2.3% had 

experienced some form of sexual violence at school or immediately outside the school gates 

(Ncontsa & Shumba, 2013). Olweus & Limber (2010) observed that in Norway, 15% of the 

school children were either involved bullies or bully victims. Elimination of school violence 

can substantially contribute towards countries’ development capacity. Efforts towards 

eliminating school violence in South Africa can therefore also contribute towards development 

capacity in South Africa (Zainal & Zainah, 2014).  

 

In the United States, a survey carried out by Wang, Iannotti, and Nansel (2009) found that 

20.8% of the school population were either victims or actual bullies. Cross et al. (2009) found 

that in Australia, just over one in every four students from Grade 4 to Grade 9 (27%) are 

involved in bullying behaviour; consequently, Australia is regarded as having the highest rate 

of bullying in the developed world (Carr-Gregg & Manocha, 2011). These percentages are 

alarming as they show the prevalence of bullying behaviour among learners despite policies 

and strategies in place to reduce bullying behaviour in schools. Carr-Gregg and Manocha 

(2011) acknowledge that in the African context studies on bullying have only recently been 

carried out, while worldwide the high prevalence rates of bullying in schools have been 



  

3 
 

documented over an extensive period of time. The new African efforts can include doing 

research that will help raise awareness of the dangers of violence in communities and schools. 

This study has focused on bullying as it is the most common form of school violence and one 

of the causes of teenage suicide (Boyes, Bowes, Cluver, Ward, & Badcock, 2014; Farmer, 

Lane, Lee, & Lambert, 2012; Hicks, Jennings, Jennings, Berry & Green, 2018). It aims to look 

at bullying behaviour and psychosomatic health among secondary school township learners. 

 

1.3 Problem Statement 

 

Globally, bullying is a prevalent concern due to its role in fomenting school violence and health 

problems (Modecki, Minchin, & Harbaugh, 2014; Zainal & Zainah, 2014). There is a growing 

consensus among researchers that bullying behaviour in school is a universal problem that can 

negatively impact individuals’ school attendance and academic performance (Glew, Ming-Yu 

Fan, & Katon, 2005). Research conducted in the Netherlands indicates that children who are 

bullies or victims of bullying are associated with an increased risk of physical and 

psychological problems (Fekkes, Pijpers, & Verloove-Vanhorick, 2005). De Wet (2005:82) 

argues that the physical consequences include “headaches, bed-wetting, loss of appetite, poor 

posture and stomach problems”. Psychologically, they are “depression, suicide tendencies and 

actual suicides, tension, fear, as well as feelings associated with posttraumatic stress, confusion, 

anxiety, anger and grief”. Rigby (2003) sums up the four main consequences as physical un-

wellness, low psychological well-being, psychological distress, and poor social adjustment. In 

South Africa in particular, violence among learners is well documented, resulting in serious 

injuries, hospitalisation of the victims, and death (Rondganger, 2016).  

 

Research in bullying among secondary school students in Kuala Lumpur, Malaysia has also 

found that bullying can affect a learner’s future life, as it can cause physical effects such as an 

individual being permanently paralysed (Zainah & Zainal, 2014), and can also result in 

individuals not being able to reach their future goals or being unattractive (Mampane & 

Bouwer, 2011). Bullying has caused a variety of social and academic problems for young 

people, beginning in primary school and continuing throughout high school (Teasley & 

Nevarez, 2016). “Social consequences of bullying include among others isolation and 

loneliness, victims have problems in mixing with other children and adults, and are/become 

very shy” (De Wet, 2005:82.) Bullying negatively impacts student’s academic performance 

and their chance to learn (Shellard & Turner, 2004), school absenteeism and disliking school 



  

4 
 

are reported as likely outcomes for victims, which eventually lead to a high rate of school 

dropouts (Rigby, 2001).  

 

Although bullying is a problem that continues even in secondary schools (Glew et al., 2005), 

less research has been conducted in secondary schools due to scholarly views that conversion 

from primary to high school is often associated with a decrease in bullying during the early 

years of adolescence (Bibou-Nakou, Asimopoulos, Hatzipemou, Soumaki, & Tsiantis, 2013). 

A study conducted in South African schools (CJPC, 2008) focusing on the relationship between 

bullying and wellness found that there was an increase in bullying: in 2008, 15.3% of learners 

experienced violence or bullying at school, whereas in 2012, the percentage was 22.2%. This 

increase indicates a worrying prevalence of bullying in schools and, therefore, the research was 

conducted to assess the influence that bullying has on the behaviour and psychosomatic health 

of learners. 

 

In particular, this study will focus on learners at the high-school level in township schools. This 

focus on township schools is due to the following reasons. Firstly, township learners are 

estimated to comprise 9 186 655 of the South African urban school population of 10 992 987 

learners (Statistics South Africa [StatsSA], 2017). Ncontsa and Shumba (2013) found that 

about 19% of the learners decide to bunk classes or drop out of school due to 

oppression/victimisation in township schools, and this is a source of worry and concern which 

needs an urgent intervention to promote a conducive and safe school learning environment. 

The Report (CJCP, 2008) notes that “a recent study from the South African Institute of Race 

Relations revealed that 23% of learners feel safe at school, implying that 77% feel unsafe”. It 

is of paramount importance to focus on the majority / most prevalent group attending high 

schools, since “eliminating school violence can make a substantial contribution to the 

development capacity of countries, widen learning opportunities to the society and improve the 

standards of living within the community” (CJCP, 2008). 

 

Previous studies have focused mainly on rural and urban private/urban learners due to a lack 

of resources, ease of access, and communication with cohorts from these institutions, without 

taking cognisance that township schools battle with multiple challenges including a lack of 

resources, under-qualified teachers, a lack of adequate sporting facilities, a lack of adequate 

support from the government, learner parents, and other stakeholders (Burton, 2008). In such 

environments, the effects of bullying, along with other detrimental behaviour, have more dire 



  

5 
 

consequences on the educational performance and social coping levels of young people. CJCP 

( 2008) argues that “bullying often heightens susceptibility to other more serious and violent 

forms of victimisation occurring at school”. A high proportion of learners who have 

experienced bullying might experience or have experienced violent crimes.  

 

1.4 Rationale 

 

It is proposed that if school environments are safe, learners will perform well because safer 

environments are known to be conducive to teaching and learning. According to Zainah and 

Zainal (2014), globally, bullying is a major problem that is difficult to eliminate in schools. 

Being bullied has become a stigma to victims in early childhood development and learning, 

carried across from the foundation phase to adulthood (Arshad, Aslam, & Tanvir, 2016. It has 

been documented as the root cause of many psychological problems found in young adults such 

as depression, anxiety, loneliness, and low self-confidence (Sesar, Simic, & Sesar, 2013). 

 

Most researchers have focused on the prevalence of bullying (Sanchez, Romero, Frantzisko, & 

Maldonado, 2016; Solberg & Olweus, 2003), although there has been substantial research 

(Menesini & Salmivalli, 2017; Swanson & Anton-Erxleben, 2016; Zych, Baldry, & Farrington, 

2017) conducted on psychological and physical symptoms of bullying in South African 

township schools. However, this present study differs from earlier ones as it differentiates 

students involved in bullying behaviour into three groups: those who bully and are themselves 

bullied, those who only bully, those who are only bullied, and those who are neither bullies nor 

bullied. This research sought to add to the existing body of literature in bullying behaviour and 

psychosomatic issues by establishing the relationship between bullying behaviour and 

psychosomatic health among township learners in South Africa. The context of South Africa, 

its history of education, apartheid and post-apartheid reconstruction efforts, and, in particular, 

violence/bullying behaviour and psychosomatic health issues at both the national and local 

levels provide fertile ground for examining these issues (Mncube & Harber, 2012). 

Furthermore, the research sought to raise awareness of the dangers of violence in communities 

and schools in a bid to ensure that schools are safe institutions for children to learn and acquire 

knowledge.  

 

 

 



  

6 
 

1.5 Aims of the Study 

 

• To investigate the forms of bullying (physical, verbal, relational/social, and 

cyberbullying) that are prevalent among township high school learners. 

• To explore gender differences in the prevalence of each of these forms of bullying 

among township high school learners. 

• To explore the association of bullying with depression and psychosomatic health 

(headache; stomach ache; backache; feeling low, irritable or in a bad temper; feeling 

nervous; difficulties getting to sleep; feeling dizzy) among township high school 

learners. 

 

 

1.6 Research Questions  

 

1. Which forms of bullying are prevalent among township high school learners? 

2. Is there a statistically significant difference between girls and boys regarding the 

prevalence of bullying in township high schools? 

3. Is there a statistically significant relationship between bullying, depression, and 

psychosomatic health among township high school learners? 

 

1.7 Research Process 

The study adopted a quantitative approach and the data collected from the respondents were 

analysed using descriptive statistics and differential statistics, specifically logistic regression 

analysis. Recruitment into the study occurred in classes of selected schools through the 

researcher explaining the study’s aims and objectives followed by the learners volunteering to 

participate. Participation was approved upon receipt of written parental consent. The 

completion of the survey was under the guidance of the researcher when it took place in the 

selected classes. Permission to conduct the study was obtained from the University of 

Witwatersrand Ethics Committee, Gauteng Provincial Department of Education and Principal 

of each school included in the research.  

 

 

 



  

7 
 

 

1.8 Outline of the Chapters 

 

Chapter One introduces the whole study including the background, problem statements, 

significance of study, aims of the research, and the research questions. Chapter Two is a 

review of the literature on bullying. The prevalence, nature, and impact of bullying and its 

effects on psychological and psychosomatic health in school children in South Africa is 

discussed in great depth in this chapter. Chapter Three provides information regarding the 

methods used in the current study. The research design is outlined, followed by an outline of 

the instruments used in the current study, information regarding the participants of the study, 

the procedure, and ethical considerations. Chapter Four presents and analyses the empirical 

findings in relation to bullying behaviour and psychosomatic health among township learners 

in Johannesburg. Chapter Five discusses the findings of the study, followed up with the 

conclusions and appropriate recommendations. Finally, study limitations and future research 

studies are discussed.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

8 
 

                                        CHAPTER 2: LITERATURE REVIEW 

 

2.1 Introduction 

 

Bullying is a problematic phenomenon faced by schools today (Boyes, Bowes, Cluver, Ward, 

& Badcock, 2014; Farmer et al., 2012; Xaba, 2014). According to Olweus (1993), bullying is 

defined as the misuse of the imbalance of power between one or more individuals, which may 

be associated with differences in gender, age, or ethnicity, with the intention of humiliating or 

hurting another. Bullying is any negative activity envisioned as injuring the next person who is 

perceived to be less psychologically or physically powerful than the perpetrator (Glew et al., 

2005; Kulig, Hall, & Kalischuk, 2008; Olweus, 2003; Rigby, 2004).  Thus, bullying is any form 

of harm or aggression motivated by a desire to dominate and aimed at hurting someone. 

However, Smith and Brian (2000) indicate that the imbalance of power and negative behaviour 

between individuals has to be repetitive in order to be considered as bullying. Bullying also 

involves social hierarchies and most bullies are opportunists as they maintain their positions 

by victimising the lower social-status learners who are perceived as easy and vulnerable targets 

(Vaillancourt, Hymel, & McDougall, 2003). 

 

Bullying can be regarded as direct or indirect. Direct bullying consists of physical bullying 

such as hitting, taking the victim’s belongings, kicking, and/or punching (Flourentzou, 2010; 

Mynatt, Heidel, & Studer, 2014; Protogerou & Flisher, 2012; Reason, Boyd, & Reason, 2016; 

Rigby, 2004). Indirect bullying involves cyberbullying and relational (also known as 

clandestine) aggression. Relational bullying is a nonphysical form of bullying whereby the 

perpetrator aims to threaten damage to relationships and also ruin the victim’s social standing 

and reputation (Dailey, Frey, & Walker, 2015; Powell & Ladd, 2010; Smith, 2004; Swart & 

Bredekamp, 2009). Bullying can also be experienced in a verbal form; this may include teasing, 

insulting, mocking, and name-calling (Flourentzou, 2010; Protogerou & Flisher, 2012; Reason 

et al., 2016; Rigby, 2004).  

 

In addition, according to Rigby (2007), bullying can be perceived as malign or non-malign. 

Malign bullying is bullying where an individual consciously seeks to harm another individual, 

while with non-malign bullying, is when the bullying is not intentional but the belief of the 

victim that the hurt was intentional (Swart & Bredekamp, 2009). what an individual does is not 



  

9 
 

of harm to anyone. Victims of bullying are categorised based on their experiences (Glew et al., 

2005). Bullies are defined as students who bully others, while bully-victims, also known as 

provocative victims, are those who bully others and are also being victimised by other bullies. 

 

2.2 Bullies 

 

Research indicates that more boys than girls are bullies (Farrington, 1993; Forero, McLellan, 

Rissel & Bauman (1999); Isernhagen, & Harris (2004); Nguyen, Bradshaw, Townsend, & Bass, 

2017). More bullies experience psychosomatic symptoms than pupils in the other bullying 

categories (bully-victims and victims). Bullies spend most of their time with friends and 

perceive school as a dreadful place to be. Glew, et al (2005) claims that bullies are frequently 

sad on most of the days they spend at school and are normally the ones who start fights. In 

addition, according to Smith (2004), most bullies have aggressive traits and come from home 

environments in which there is less affection, low parental monitoring, and more violence. 

Bullies perceive their attacks on others and aggressive behaviour in general as a mechanism to 

gain influence and control. A study conducted in Brazil found bullying behaviour to be more 

prevalent among boys than girls, boys being more involved as both bullying perpetrators and 

bully-victims than girls. Some studies reveal girls to be more victimised than boys, while other 

studies reported opposite results, and some show that the rates of victimisation are equal 

between the two genders. For the Brazilian study, girls were not found to be more frequent 

victims than boys, although a difference was found between the adolescent girls’ subgroup 

when compared to the younger children subgroup, in which girls were more frequently victims 

than boys. The results of a 40-country cross-national study showed consistency in finding boys 

to be exhibiting a higher level of aggressive behaviour in all countries. Rates of victimisation 

were higher among girls in 29 of these 40 countries (Isolan, Salum, Osowski, Zottis and 

Manfro, 2013) 

 

Research shows that there are different types of bullies. Firstly, some are anxious or are both 

anxious and aggressive as they have low self-esteem. Anxious bullies have few friends and 

they are often emotionally unstable. Secondly, some passive bullies are not aggressive but 

easily dominated as they are empathic to others and somehow feel guilty after a bullying 

incident (De Wet, 2005). It has been reported that bullies have a high possibility of becoming 

criminals in future (Glew, Rivera, & Feudtner, 2000; Olweus, 1993), experience difficulty in 

building personal relationships (Hugh Jones & Smith, 1999), and later in adulthood eventually 



  

10 
 

have problems with substance abuse (Kaltiala-Heino, Rimpelä, Rantanan & Rimpelä, 2000). It 

has also been found that school bullying is negatively associated with several school-related 

issues, including adjustment, bonding, and perceived climate (Dussich & Maekoya, 2007).  

 

2.3 Victims 

 

Research shows that more boys than girls are victims of bullying. Victims of bullying feel 

lonely at school as other learners do not want to spend time with them. This can result in them 

hating school and missing days without their parents’ permission (Forero, McLellan, Rissel, & 

Bauman, 1999). In addition, according Glew et al. (2005), victims usually are learners who 

struggle with their schoolwork. They dislike school as they perceive it as an unsafe place to be. 

Studies show that victims of bullying are more likely to be unhappy and insecure children and 

they are very anxious, with low self-esteem (Powell & Ladd, 2010).  

 

There are two types of bully victims, namely provocative victims and passive victims (Powell 

& Ladd, 2010). Provocative victims are also known as bully-victims, as they partake in both 

sides of bullying, playing both the roles of bullies and victims (Powell & Ladd, 2010). Those 

learners who bully and are bullied by others are likely to be boys and they also experience 

psychosomatic symptoms. Bully-victims like being at school (Forero, McLellan, Rissel, & 

Bauman, 1999). Glew et al. (2005) add to Forero et al.’s (1999) argument that bully-victims 

struggle academically and are more likely to be boys than girls. In contrast to Ferero et al., 

however, Glew et al. (2005) noted that bully-victims feel unsafe and they feel that they do not 

belong at school.  

 

Powell and Ladd (2010) argue that provocative victims are physical bullies as they act with the 

intention of reprisal and they do not seek power; however, they lack close friends which results 

in them being lonelier and more insecure than those individuals who are only bullies (Powell 

& Ladd, 2010). Bully-victims likely experience continuous social, physical, and psychological 

effects and tend also to suffer from being victimised later in life (Ando, Asakura & Simons-

Morton, 2005; Banzai,1995). The research shows that bully-victims in schools are likely to be 

continuously vulnerable and are therefore also likely to be bullied later in life, as adults at their 

workplaces (Smith, Singer, Hoel, & Cooper, 2003). Passive victims are characterised as 

individuals who usually suffer from anxiety, low physical strength, and/or depression; have 

poor social skills; are lonely and quiet; as well as having low self-esteem, making them insecure 



  

11 
 

(Powell & Ladd, 2010; Smith, 2004). Research has suggested that bullying behaviours of both 

sorts and victimisation are differentially associated with various family-level factors (De la Rue 

& Espelage, 2014).  

 

A study conducted in  schools in the south-east of the United States, focusing on the 

relationship between bullying and wellness, showed that the most common type of bullying 

experienced by students is verbal, as 37.9% of students indicated having experienced verbal 

bullying, 22.3% relational bullying, 16.4% indirect bullying, 11.9% physical bullying, and 

9.1% cyberbullying (Mynatt et al., 2014). These figures support the assertion that school 

bullying is highly prevalent, and therefore help to motivate this research to reveal the influence 

that bullying has on the mental or psychological and psychosomatic health of learners in 

township high schools. 

 

2.4 Gender Differences and Bullying 

 

Researchers have perceived bullying to be experienced differently across different genders 

(Due et al., 2005; Espelage, Mebane, & Swearer, 2004). According to Smith (2014), gender 

differences show a discrepancy in the type of bullying that males and females experience. 

Globally, most studies revealed that males are more likely to experience physical forms of 

victimisation while females are more likely to be involved in either verbal or relational bullying 

(Anaya, 2015; Greeff & Grobler, 2008; Smith, 2014). However, a study on the relationship 

between bullying and wellness (Mynett et al., 2014) reported that 46.5% of male students were 

involved in verbal bullying, followed by 17.9% in relational bullying, 17.8% in physical 

bullying, 13.2% in indirect bullying, and 3.6% in cyberbullying. For female students, 32.8% 

indicated that they use verbal bullying, 25.5% relational bullying, 18.7% indirect bullying, 

12.7% cyberbullying, and 8.1% physical bullying. These incidents tended to occur weekly and, 

in any location, (Mynett et al., 2014).  The figure supports Smith’s (2014) conclusion that males 

are more involved in physical bullying than girls (17.8% compared to 8.1% in the Mynatt 

study). In contrast to Smith (2014), though, Mynatt et al. (2014) also indicated that males are 

more involved in verbal bullying as they scored 13.7% more than females. However, a study 

conducted in Hong Kong found that there is no gender difference for verbal bullying or social 

exclusion (Smith, 2014). 

 



  

12 
 

It is believed that the genders have different risk profiles. Most boy victims are at risk of anxiety 

which results in them attempting suicide, while aggressive boys are at risk of personality 

disorders. On the other hand, for girl victims suicide attempts and suicide was predicted even 

though the problem levels were controlled in the early stages of life (Isernhagen & Harris, 

2004; Undheim & Sund, 2010). 

 

2.5 Prevalence of Bullying/Victimisation 

 

According to Rassool (2002), research conducted in a South African Catholic school in 2001 

found that more than 90% of the students reported that they were bullied. which indicates that 

prevalence can be extremely high. As previously indicated, boy learners are more likely to 

bully their peers than girls are as most boy bullies tend to be physically stronger than their 

chosen victims. In addition, the American study that addressed influences of race on bullying 

found that white or Hispanic youth reported more victimisation than black students. Children 

with learning difficulties, emotional, or behavioural problems were more likely to be involved 

in bullying than their peers (De la Rue & Espelage, 2014). However, many talented children 

and adolescents also fall victim to bullying, which prevents them from achieving their fullest 

academic potential (De la Rue & Espelage, 2014).  

 

Research indicates that teachers often report a lower prevalence of bullying than students do 

(Burton, 2008). Studies conducted by Burton (2008; 2012) in South African schools show that 

bullying behaviour and school violence increased from 15.3% in 2008 to 22.2% in 2012. The 

2012 report notes that in that year alone, 12.2% of the sample reported being threatened with 

violence, 6.3% were assaulted, 4.7% were sexually assaulted/raped, and 4.5% were robbed at 

school. In addition, in Katlehong, east of Johannesburg, a 13-year-old girl was sexually gang-

assaulted by three 14-year-old boys. The perpetrators used their cell phones to film the violent 

assault, which is an example of the increasingly common use of technology in the context of 

violence among in high school pupils (Anonymous, 2010). 

 

2.6 Traditional Bullying and Cyberbullying 

 

Cyberbullying is defined as bullying done by means of technological devices through 

electronic avenues such as social networks, constant messaging, websites, and emails 

(Kowalski & Limber, 2013; Olweus, 2012; Powell & Ladd, 2010). Sánchez, Romero, Navarro-



  

13 
 

Zaragoza, Ruiz-Cabello, Frantzisko and Maldonado (2016), note that cyberbullying is common 

among pupils aged 10 to 17 years as they use the internet regularly. Research shows that 

adolescents experience cyberbullying more than the university population, ranging from 9 to 

20% of the samples studied (Sánchez et al., 2016). Cyberbullying can negatively impact on 

students’ education, which may result in them carrying weapons to school or cause students to 

skip school (Powell & Ladd, 2010). Olweus’s (2012) study challenged claims that traditional 

forms of bullying were common than cyberbullying in schools. In addition, Menesini (2012) 

believes that cyberbullying is underreported as there is a lack of clarity on its definition, and 

adds that in order to determine the prevalence of cyberbullying accurately, it is essential that 

cyberbullying occurrence be studied separately and given its own particular forms of measure. 

 

According to Modecki, Minchin and Harbaugh (2014), and Modecki, et al’s (2014) study 

conducted in Australia, the prevalence rate of cyberbullying and traditional bullying differ. The 

prevalence of traditional bullying involvement is 35% while cyberbullying is 15%. A study 

conducted among high school learners in Spain found an overlap between cyberbullying 

victimisation and traditional bullying (Sánchez et al., 2016). These studies show no 

consistency, however, as cyberbullying and traditional bullying are experienced differently 

according to different countries.  

 

2.7 Consequences of Bullying 

 

As previously mentioned, bullying has long-lasting consequences for the individual who has 

been or is being bullied; such effects include psychological and emotional trauma. Holt, Chee, 

Ng, and Bossler (2013) state that recurring bullying of an individual makes them more 

susceptible to social, physical, and mental health consequences in the long run. Bullied victims 

have the misfortune of developing a propensity for these problems which, in turn, affects how 

they perform in their academic work and the frequency of school attendance (Holt et al., 2013; 

Butler, Kift, & Campbell, 2009). In this instance, as evidenced, bullying can play a huge role 

in how the victimised individuals see themselves as compared to their peers who are not being 

bullied, i.e., feeling inadequate to deal with the challenges of attending school or not seeing the 

benefit of going to school only to be bullied. 

 

Longitudinal studies have been conducted that support the hypothesis that one of the effects of 

being bullied is psychological trauma, providing evidence of the long-term effects on the 



  

14 
 

psychological well-being of the bullied victim into adulthood (Rigby, 2003). Being bullied 

“permanently damages both the psyche and education of the child” (Laas & Boezaart, 

2014:2667). 

 

2.8 Health Problems 

 

Bullying has been associated with psychological, behavioural, and psychosomatic symptoms 

such as high levels of anxiety, depression, loneliness, suicidal behaviour, nightmares, and low 

self-confidence (De la Rue & Espelage, 2014; Sesar et al., 2013). The term “psychosomatic 

symptoms” refers to both physical and psychological complaints of which the causes are mental 

and emotional rather than physical. The physical symptoms include headaches, stomach-aches, 

backaches, difficulty getting to sleep and feeling dizzy. The list of psychological complaints 

includes depression, irritability or bad temper, nervousness, and feeling low (Seixas, Coelho, 

& Nicholas-Fischer, 2013). In a school in Benoni, Johannesburg, the alleged bullying of a 13-

year-old boy is said to have resulted in his death as he suffered from severe headaches (Alfreds, 

2017). According to Ayenibiowo and Akinbode (2011), chronically victimised students may 

experience poor self-esteem, depression, and other mental health problems including eating 

disorders (anorexia and bulimia nervosa) as well as schizophrenia. Studies show that there is a 

relationship between bullying and psychosomatic symptoms (Sesar et al., 2013).  

 

2.9 The South African Legislation on Bullying 

 

In many countries, laws are put in place to ensure that victims of bullying receive justice and 

that they are accordingly looked after. In such cases, their best interests are at the core of 

making and implementing laws in and around the school environment, especially. Various 

studies have provided evidence that a great deal of bullying happens in the school environment, 

thus calling for a critical intervention to ensure that the rights of the offender and victim are 

not being violated (Laas & Boezaart, 2014). The South African legislative framework on 

bullying looks at the rights which may be or have been violated in the bullying process. 

According to the Constitution of the Republic of South Africa, bullying involves the violation 

of individuals’ constitutional rights (Laas & Boezaart, 2014). 

 

Among the rights that are being violated when bullying takes place is the victims’ right to 

equality. According to the Constitution of South Africa, “no individual may discriminate 



  

15 
 

against another person based on race, gender, sex, pregnancy, marital status, ethnic or social 

origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture, language 

and birth” (Juta Law, 2017). Another right is human dignity; when someone is being bullied, 

it is clear that individual’s dignity is being infringed upon as they are seen as less human than 

they are. The motive of a bully is to intimidate, wield power, and humiliate; thus, the victim’s 

dignity and even self-worth are negatively impacted (Laas & Boezaart, 2014). Section 12 of 

the Constitution focuses on protection against various kinds of violence that may negatively 

impact on an individual. It therefore protects learners from acts of bullying in any form. 

“Section 12 of the Constitution intricately ties into the bullying phenomenon, as the right to the 

freedom and security of the person protects everything that bullying transgresses” (Laas & 

Boezaart, 2014:2676).  

 

Bullying also infringes on the victim’s right to access to education. Victims may feel the need 

to drop out of school and may not perform schoolwork adequately as they are being bullied.  

 

South Africa’s national legislation on bullying is built on the constitutional framework 

discussed above. In the school context, four acts are central to analysing the legal situation, 

namely the South African Schools Act 84 of 1996, the Children’s Act 38 of 2005, the Child 

Justice Act 75 of 2008, and the Protection from Harassment Act 17 of 2011 (Laas & Boezaart, 

2014). 

 

The following provides a detailed explanation of each act, with special emphasis on the topic 

of bullying in schools. The first act to be discussed is the South African Schools Act 84 of 1996 

(SASA), which dictates that a governing body of any public school should adopt a code of 

conduct for their learners after consultation with relevant stakeholders. The code must be 

phrased in a way that it enforces discipline and creates a purposeful school learning 

environment. It must outline positive behaviour and, in turn, how misbehaviour will be handled 

and how and when disciplinary measures will be executed (Juta Law, 2017). “A code of 

conduct provides the regulatory framework within which a safe, secure, disciplined and 

purposive learning environment may be established. It must also improve and maintain the 

quality of the educating process” (Laas & Boezaart, 2014:2679). 

 

According to Ngqakamba (2019) In South Africa, children as young as 10 years may be 

criminally charged for bullying behaviour if they are found to be guilty beyond a reasonable 



  

16 
 

doubt that they were conscious of their actions and consequences. Even though bullying itself 

is not categorised as a criminal offence, behaviours associated with it could be classified as 

crimes in terms of the Criminal Procedure Act; for example, crime injuria, assault, extortion, 

and attempted murder. The whole procedure of children being accountable for their bullying 

tendencies, however, is more aimed at restoration than punishment, in that the state wants to 

assist children found guilty of crimes instead of punishing them (Ngqakamba, 2019). 

 

South African laws may not yet recognise bullying as a crime, but there are laws in place that 

ensure children’s rights are being protected regardless (Law for All, 2017). In the Children’s 

Act 38 of 2005, the emphasis is protecting children against neglect and abuse; it is also aimed 

at eradicating bullying in schools, allowing every bullied victim an opportunity to present their 

case in a court of law. The aim of this law is not to punish the bully but rather to maintain the 

balance that was thrown off when one child was victimised. This act, in conjunction with SASA 

Act, aims for restorative justice through specific programmes and processes (Law for All, 

2017). 

 

The child Justice Act 75 of 2008 calls for restorative justice but does not neglect cases of 

bullying that includes criminal intent. The law calls for a separate justice system for children 

who might be involved in the bullying process. The act created three categories that divide 

where a person falls under which category. The first category is: children below 10 years of 

age, the second one is children 10 years and older but younger than 18 and the third one is 

young people 18 years of age and older but under 21 years. As in this discussion bullying is 

discussed in the context of school, this usually means more often the perpetrator will be a child 

or a learner-on-learner bullying, thus this type of context will fall under the first two mentioned 

categories. Even before the child is considered to have a criminal intent, they assess whether 

the child has criminal capacity or not (Law for all, 2017). 

 

The protection from harassment Act of 2011 is the last legal framework which is used to 

safeguard the rights of the children.  This law states that individuals or victims of bullies are 

by law have the right to apply for a protection order against their perpetrators, if they have 

suffered or have been bullied by their fellow school mates. According to law for all blogs, 

under section 2(4) of the protection from Harassment Act states that any child who is a victim 

of being bullied can apply for a protection order, regardless of the statement that a parent or 

guardian should accompany the child to open the protection order (Law for all. 2017). 



  

17 
 

 

2.10 The Constructs between Bullying and Psychosomatic Health 

 

School bullying is a phenomenon that worries health professionals, psychologists, teachers and 

families in many countries around the world. Researchers’ attention towards bullying has 

increased in particular the relation between bullying and psychosomatic problems. Gini and 

Pozzoli (2013) have shown that bullies are likely to display negative and antisocial behaviour, 

for example, truancy, delinquency, and substance abuse, during adolescence and are at risk for 

psychiatric disorders. Furthermore, Gini and Pozzoli (2013) found that their victims were prone 

to suffer from a variety of psychosomatic and behavioural problems, for example tiredness, 

nervousness, sleeping problems and dizziness. Worthy of note is that tiredness and sleepiness 

are regarded as symptoms of depression.  

 

2.11 Theoretical Framework 

The prevalence of bullying and its effects on psychosomatic health appear to warrant an 

extensive investigation. A number of theories have been put forward to explain the occurrence 

of bullying and its effects among school learners around the globe. Below is a review of the 

Social Dominance Theory which sought to explain the prevalence of bullying and its effects 

among school learners 

 

2.11.1 Social Dominance Theory 

The social dominance theory (SDT) is based on the belief that human systems and societies are 

structured according to social group-based hierarchies. These hierarchies include a dominant 

and submissive group (Goodboy, Martin, & Rittenour, 2016). The difference between them is 

related to the accomplishment of positive social status, which is measured according to the 

accumulation of desirable social and financial resources. The SDT explores the notion of 

individual social power that is gained through membership in a socially constructed group 

(Goodboy et al., 2016). According to Sidanius and Pratto (1999), SDT postulates that 

communities contain status hierarchies, as some groups of people in the community are more 

privileged than others. As a result, inequalities exist between these groups based on race, age, 

gender, and imbalance of power (Sidanius et al., 1994). Although there are intrapersonal 

characteristics that contribute towards children being at risk for bullying, the SDT’s interest 



  

18 
 

lies within the social hierarchical construct. The concept of social hierarchies provides a 

conceptual framework that explains bullying (Romm, 2013). 

 

According to the SDT, bullying is used as a tool to maintain social dominance within these 

social hierarchies. Dominance is not an end in itself but a means to get prioritised access to 

resources that are valued for the group. People use bullying behaviours such as aggressive 

physical, emotional, and social bullying in order to gain and maintain dominant status within 

the hierarchy (Goodboy et al., 2016). The theory does not indicate that this behaviour is 

prevalent because children are innately evil, but rather that this behaviour is adopted in order 

to survive and maintain ranking within school classes and peer groups. In order to be 

successful, children need to be socially skilled. Looking at the lower end of the dominance 

hierarchy, children who are aggressive tend to lack social skills and these become bully-victims 

or provocative/aggressive victims (Romm, 2013). Therefore, their primary intention of 

bullying is not to cause harm in itself but rather instrumental in a struggle to acquire or maintain 

status, and is used in a calculated way. Aggression is often a tool. In particular, individuals use 

aggression in order to disrupt the hierarchy within new groups. Therefore, according to the 

SDT, bullying is usually used as a recourse to increase social dominance status, and once this 

is achieved, the bullying decreases (Goodboy et al., 2016). 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

19 
 

 

CHAPTER 3: RESEARCH METHODOLOGY 

 

3.1 Introduction 

This chapter outlines the research methodology that was employed in the study. The contents 

of this chapter include: the research paradigm, research design, participants, instruments, 

procedure, data analysis and ethical considerations of the study. 

 

3.2 Theoretical Framework 

The theoretical framework influences the way knowledge is studied and interpreted (Mertens, 

2005). The theoretical framework referred to in Section 2.11, Social Dominance Theory, has 

been used to decide on the variables to be included in the present study, in particular the 

independent variables bullying and gender, and the dependent variable, psychosomatic health 

problems. According to the SDT, various types of bullying per gender happened more among 

boys than girls while psychosomatic health challenges happened more among girls than boys. 

According to the SDT, bullying is usually used as a recourse to increase social dominance 

status.  

 

3.3 Research Paradigm 

The choice of paradigm sets down the intent, motivation and expectations for the research. 

Mackenzie and Knipe (2006) argue that without nominating a paradigm as the first step, there 

is no basis for subsequent choices regarding methodology, literature or research design. A 

number of paradigms have been discussed in the literature, such as: positivist, constructivist, 

interpretivist, transformative, emancipatory, critical, pragmatic and deconstructivist 

(Mackenzie & Knipe 2006). This research will adopt a positivist paradigm. Positivism is 

referred to as scientific method or science research that aims to test a theory or describe an 

experience “through observation and measurement in order to predict and control forces that 

surround us” (O’Leary, 2004:5). Silverman (2000) noted that positivist paradigm tends 

predominantly to use quantitative approaches (methods) to data collection and analysis. This 

research seeks to investigate the association between bullying and psychosomatic health and 

in attempting to find out if any association or relationship exists between them, a quantitative 

approach has been adopted. The quantitative research method is guided by the positivist 

paradigm in adopting certain approaches that require measurement. 



  

20 
 

 

3.4 Research Design  

Quantitative research is categorised as an objective positivist search to get singular truths that 

depend on the variables, hypotheses, and statistics; it is generally large scale, but without much 

depth (O’Leary, 2004). The research design adopted is a cross-sectional survey, as the sampling 

of participants was conducted at a single point in time rather than over a period of time. 

  

3.5 Participants 

 

The target population for this study was high school learners from Orlando West high schools 

in Soweto, Johannesburg. Three hundred learners were sampled. A convenience sampling 

technique was used to select respondents. Etikan, Musa, and Alkassim (2016:2) describe 

convenience sampling as “a type of non-probability or non-random sampling where members 

of the target population that meet certain practical criteria, such as easy accessibility, 

geographical proximity, availability at a given time, or the willingness to participate are 

included for the purpose of the study”. Criteria for inclusion were that the participants should 

be learners aged 14 to 17 years who were grade 8 to 11 in Orlando West high schools, in 

Soweto, a township. Soweto is an urban settlement adjoining Johannesburg, in the province of 

Gauteng, South Africa. It is populated by approximately 1.3 million people (Loots, 2008). The 

2012 National School Violence study found that there is a high rate of bullying and violence 

in Gauteng township schools (Burton & Leoschut, 2013) which made the Soweto schools 

chosen ideal for this research. 

 

Previous studies drew their conclusions about the association between psychosomatic 

symptoms and bullying using smaller samples; as a result, the researcher of this paper aimed 

to focus on a more generous sample of 300 participants to investigate the correlation between 

psychosomatic health and bullying among secondary school township learners. Research 

indicates that there is a relationship between bullying involvement and parental maltreatment 

(De la Rue & Espelage, 2007). According to Rezapour, Soori, Nezam Tabar, and Khanjani 

(2020), previous studies focusing on the prevalence of psychosomatic symptoms in European 

school-aged children found that at least one psychosomatic symptom was experienced by 

45.7% of the children involved in bullying. 

 

 



  

21 
 

3.6 Instruments  

 

This research study used a demographic questionnaire (see Appendix A), the Revised Olweus 

Bully/Victim Questionnaire (Appendix B), Cyberbullying Behaviours and Victimisation 

Experiences Measure (see Appendix C), General Health Questionnaire (Appendix D), and 

questionnaires on Psychosomatic Problems (Appendix E), as well as Cyberbullying & Bullying 

(Appendix F). All these questionnaires were conducted in English as it is the medium of 

instruction in the South African education system. However, translation into and explanation 

through any local language was provided when administering the questionnaires. 

 

3.6.1 Demographic Questionnaire 

Kitchenham and Pfleeger (2002) point out the significance of having a demographic 

questionnaire as it provides information that describes the participant based on his/her age, 

gender, ethnicity, home language, school type, and grade (refer to Appendix A). These 

demographic aspects were useful for descriptive purposes as well as analysing gender 

differences in the prevalence of each form of bullying among the participating schools. 

 

3.6.2 The Revised Olweus Bully/Victim Questionnaire (R-OBVQ)  

The Revised Olweus Bully/Victim Questionnaire (R-OBVQ) is a self-reported measure that is 

commonly used to assess individuals who experience bullying and being bullied by others 

(Solberg & Olweus, 2003). Although this measure consists of 40 questions, for the current 

study, only 23 items were used. The researcher selected those which specifically focused on 

the nature of the research. The 23 questions focused on issues connected with and exposure to 

various forms of bullying, and attitudes as well as individuals’ reactions to bullying (Violence 

Institute of New Jersey, 2002). According to Rivera (2014), the OBVQ measure is found to 

have satisfactory internal reliability with Cronbach’s alpha values greater than 0.80. This 

measure demonstrates suitable evidence of construct validity when assessing associated 

variables and relations between the dimensions of “being victimised” (Bendixen & Olweus, 

1999; Solberg & Olweus, 2003). A study conducted in Bloemfontein, South Africa used the R-

OBVQ with specific reference to its applicability within the South African context in order to 

acquire descriptive information about the prevalence of bullying in schools (Greeff & Grobler, 

2008).  

 

 



  

22 
 

3.6.3 Cyberbullying Behaviours and Victimisation Experiences Measure 

The Cyberbullying Behaviours and Victimisation Experiences Measure is an instrument that 

assesses individual victimisation experience and cyberbullying behaviour (refer to Appendix 

C). This measure has five questions that are assessed using a 5-point Likert scale; the scale 

ranges from 0 = never to 4 = every day (Holfeld & Leadbeater, 2015a). According to Holfeld 

and Leadbeater (2015b), convergent validity was found for victimisation and bullying.  

 

3.6.4 General Health Questionnaire (GHQ-12)  

The General Health Questionnaire (GHQ-12) was developed by Goldberg in the 1970s 

(Appendix D). The instrument is a brief self-report measure which aims to assess psychiatric 

disorders in non-clinical settings (Goldberg & Williams, 1988). This measure consists of 12 

items which contain six ‘positive’ and ‘negative’ items that focus on the past few weeks (Hu, 

Steward-Brown, Twigg, & Weich, 2007). The items are assessed using a four-point scale in 

which a high score on negative items indicates greater distress and/or difficulty and a high 

score on positively worded items indicates endorsement. The validity of the GHQ-12 was 

assessed using convergent validity and internal consistency was used to test reliability of the 

questionnaire using Cronbach’s alpha coefficient (Montazeri, Harirchi, Shariati, & 

Garmaroudi, 2003).  

 

3.6.5 The Psychosomatic Problems (PSP) scale 

According to Hagquist (2009), the PSP scale is used to assess psychosomatic health (Appendix 

E). This measure is constructed through the summation of raw scores of learners’ responses, 

consisting of eight items using a five-point scale. The scale ranges from never to always 

(Hagquist, 2009). The study conducted in Sweden used the PSP scale to describe trends in 

psychosomatic health problems among adolescents, with a focus on gender differences; the 

scoring of the psychometric analysis used the Rasch model 11-13 (Hagquist, 2009). This 

instrument was piloted. 

 

3.6.6 Cyberbullying and Bullying  

The Cyberbullying and Bullying measure assesses the frequency of bullying using a five-point 

scale. The scale ranges from never to several times a week. This instrument consists of four 

items, exploring how an individual experiences cyberbullying during a particular period of 

time. This instrument was piloted. 

 



  

23 
 

 

3.7 Procedure  

 

The researcher ensured that ethical clearance was attained from the Humanities Research 

Ethics Committee (Non-medical) in order for the study to proceed and be conducted. The 

researcher then requested permission from the Department of Education and the school heads 

in Orlando West secondary schools. Once permission was granted by the Department of 

Education and the school heads, parents and learners were given information sheets and 

consent forms to give parental permission. Learners were approached after school, when the 

researcher introduced them to the research study and stated the potential benefits and risks of 

participation in it.  

 

Learners were informed about the guaranteed anonymity and confidentiality and notified that 

their participation is voluntary. They were not asked to place any form of identification on the 

questionnaires, as participant numbers but no names were used on the questionnaires and 

reports. The researcher administered the questionnaires to ensure that any questions or 

problems raised were responded to. A sample of 300 learners who were willing to participate 

in the study was given a set of self-report questionnaires. 

 

The learners were given about 45 to 60 minutes to complete answering the research 

questionnaires. For those participants who experienced any discomfort or emotional distress 

during participation, a de-briefing was presented by a qualified clinical psychologist 

(Nokulunga Nene) and contact details of free counselling services were provided to participants 

as well as the contact details of the researcher for those participants who were interested in the 

outcome of the study.  

 

Immediately after the questionnaires were completed, raw data was placed in a sealed box, 

after which it was transferred to a locked cabinet. Thereafter, raw data was converted into 

digital format using the Microsoft Excel software. This was done to ensure a de-identifiable 

dataset in a password-locked computer. Thereafter, the data was analysed using the statistical 

software SPSS (Statistical Package for the Social Sciences) version 24. The results of the study 

were reported in the form of a master’s research report. 

 

 



  

24 
 

3.8 Data Analysis 

 

This research aimed to look at bullying behaviour and psychosomatic health among secondary 

school township learners. It specifically looked at the prevalence of forms of bullying, gender 

differences in the prevalence of each form of bullying, and psychosomatic health. Descriptive 

statistics and inferential statistics were derived. Descriptive statistics were used to explore the 

frequencies and percentages relating to the prevalence of bullying. Inferential statistics were 

used to explore whether there are significant differences between bullying and psychosomatic 

health. Pearson’s product moment correlation was used, provided assumptions of parametric 

tests were satisfied. If not satisfied, Spearman’s correlation was used. SPSS software was used 

to analyse the data. 

 

3.9 Ethical Considerations 

 

The researcher obtained ethical clearance from the Research Ethics Committee (Non-medical) 

before conducting this research study (protocol number: MEDP/18/006 IH). The Gauteng 

Department of Education was informed of the study in order for the researcher to receive 

permission to conduct it at the Johannesburg West District 12 high schools in Orlando West. 

The heads of the high schools chosen were informed through an information sheet and consent 

form in order for the researcher to receive the requisite permission from them. The participants 

were informed that the research is voluntary and for academic purposes. 

 

The researcher ensured that all questionnaires were confidential and anonymous. Participants 

were given informed consent; information concerning informed consent was clarified and 

disclosed before participants undertook the research questionnaire. The researcher asked for 

permission from the participants’ parents as the participants are minors and a vulnerable group 

to work with (see appendix). The researcher also asked the participants to give assent (see 

appendix) and the participants were informed that they could withdraw from answering the 

questionnaire should they feel the need to. Contact information of both the researcher and 

researcher’s supervisor were provided to the participants and after the questionnaire was 

completed a qualified clinical psychologist (Nokulunga Nene) provided a briefing addressing 

bullying. She also assisted when any problem arose requiring her attention. Participants were 

provided with contact details of toll-free counselling services, which are as follows:  

• The South African Depression and Anxiety Group: 0800 567 567 / 011 262 6396 



  

25 
 

• Lifeline South Africa: 0861 322 322. 

 

The dataset from participants was kept on a password-protected computer. Only the researcher 

and supervisor were able to view the raw and digital dataset collected. The study results were 

reported in the form of a master’s research report which is accessible on the online Wits 

research repository and also at the main library of the University of the Witwatersrand.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



  

26 
 

CHAPTER 4: RESULTS 

 

4.1 Introduction 

 

This chapter aims to discuss the results obtained from the analysis of data collected. The 

research questions were used as guidance to present the results obtained from the data. The 

investigation thus addressed the prevalence of psychosomatic health conditions, forms of 

bullying, and gender differences in bullying (physical, verbal, relational/social, and 

cyberbullying), as well as determining whether there is a relationship between bullying and 

psychosomatic health among township high school learners in Orlando West. The investigation 

results are presented below as univariate, bivariate, and multivariate analyses of the data 

through descriptive statistics as well as logistic regression. Univariate analysis is a method for 

analysing data on a single variable at a time, thereby observing only one aspect of the 

phenomenon. Univariate analysis explores age and/or sex as single-variable data that can be 

illustrated on pie charts or histograms. Bivariate analysis is used to explore data of only two 

variables under consideration. The bivariate technique reveals a relationship between two 

variables, for instance, level of education and income. Multivariate analysis is a simultaneous 

study of several variables. It is more informative than univariate analysis. Multivariate analysis 

explores three or more variables together to understand their various relationships.  

 

4.2 Descriptive Outcome 

 

Univariate analysis 

Age was found not to be normally distributed upon testing for normality, and showed a median 

of 16 years and an inter-quartile range of three years. In terms of gender, 65% of the sample 

were males while 35% were females. 



  

27 
 

 

 

Figure 4.1. Gender distribution across sample. 

 

Looking at the educational level of the sampled participants, the highest proportion (42%) were 

grade ten learners, 32% were in grade eight, 15% were in grade eleven, and grade nine learners 

made up the lowest proportion at 11%.  

 

 

 

 

 

 

 

 

 

 

 

 

Figure 4.2. Educational level. 

 

The most represented home language was isiZulu, spoken by 60% of learners, followed by 

Sesotho and Setswana at 11% each, as opposed to the least represented languages, at 1% each 

for English and Ndebele. 

Males
65%

Females
35%

Gender distribution across sample

Grade 10 

42% 

Grade 11 

15% 

Grade 8 

32% 

Grade 9 

11% 

Educational Level of Participants 



  

28 
 

 

 

Figure 4.3. Participants’ home languages. 

 

A quarter of the sample, 25%, self-identified as bullies, 22% were bullied and 8% were bully-

victims. 

 

 

Figure 4.4. Levels of bullying. 

  

 

 

0

5

10

15

20

25

Bullied Bully Bullied-Bully

22

25

8

Pe
rc

en
ta

ge
 (%

)

Levels of Bullying Behaviour among Adolescents living in Townships, 
Gauteng, South Africa 2018



  

29 
 

Psychosomatic health 

The prevalence of psychosomatic health conditions is shown below in Figure 4.5. Results 

indicate that 46% of learners suffered from psychosomatic health conditions, while 54% of the 

sampled population did not.  Figures 4.6 to 4.13 show the results for individual types of 

psychosomatic health conditions (i.e. difficulty concentrating, difficulty sleeping, headache, 

stomach ache, feeling tense, loss of appetite, feeling sad,, and feeling dizzy). 

 

 

Figure 4.5. The prevalence of psychosomatic health conditions in township schools. 

 

Figure 4.6 demonstrates that 51% of learners indicated that they never had difficulties in 

concentrating, while 29% of the sample indicated that they sometimes did, 11% that they 

seldom had this problem, 6% said ‘often’, and 3% indicated that they always experienced 

difficulties in concentrating. 

 

Never

51%
Seldom

11%

Sometimes
29%

Often
6%

Always
3%

Difficulty Concentrating



  

30 
 

 

Figure 4.6. Difficulty concentrating. 

 

Asked whether they had difficulty in sleeping, 55% of the sampled learners said “never”, 26% 

said “always”, 9% said “seldom”, 5% each said “often” or “always”. 

 

 

Figure 4.7. Difficulty sleeping. 

 

Headaches would seem to be fairly common, with the largest group, at 39%, stating that they 

sometimes had headaches, and 36% reporting that they never did. For the rest, 10% said 

“seldom”, 8% said “often”, and 7% said “always”. 

 

 

Never
55%Seldom

9%

Sometimes
26%

Often
5%

Always
5%Difficulty Sleeping

Never
36%

Seldom
10%

Sometimes
39%

Often
8%

Always
7%

Headaches



  

31 
 

Figure 4.8. Headaches. 

 

As figure 4.9 shows, results indicate that 41% of the sampled learners never experienced 

stomach ache, while 34% had stomach ache sometimes, 13% seldom, 9% often, and 3% 

always. 

 

 

Figure 4.9. Stomach ache. 

 

Feeling tense was a less frequent occurrence, with only 5% and 4% of the sample respectively 

reporting that they often or always felt tense. The majority, at 55%, stated that they never felt 

tense and 22% of the sample reported that they sometimes did. 

 

 

Never
41%

Seldom
13%

Sometimes
34%

Often
9%

Always
3%

Stomach Ache

Never
55%Seldom

14%

Sometimes
22%

Often
5%

Always
4%

Felt Tense



  

32 
 

Figure 4.10. Feeling tense. 

 

The results illustrated below show that 38% of learners indicated that they had never 

experienced loss of appetite, while 32% sometimes did. Further, 11% often had a loss of 

appetite and 5% of the sample reported that they always experienced loss of appetite. 

 

 

Figure 4.11. Loss of appetite. 

 

Figure 4.12 indicates the levels of sadness reported. As can be seen, 35% of the sampled 

learners reported that they never felt sad, but slightly more, 37%, indicated that they felt sad 

sometimes, while 9% often and 8% always felt sad. 

 

 

Never
38%

Seldom
14%

Sometimes
32%

Often
11%

Always
5%

Loss of Appetite

Never
35%

Seldom
11%

Sometimes
37%

Often
9%

Always
8%

Felt Sad



  

33 
 

Figure 4.12. Feeling sad. 

 

Results indicated that 44% of the sampled learners never felt dizzy. At the other end of the 

scale, 8% often and 3% always experienced dizziness. The remainder sometimes (27%) or 

seldom (13%) did. 

 

 

Figure 4.13. Feeling dizzy. 

 

4.3 Inferential Outcome 

 

Bivariate analysis 

To establish gender differences in the prevalence of bullying types, a chi-square test was used 

to see whether the levels of bullying were the same or different among boys and girls. As seen 

in Table 4.1 below, girls and boys reported similar levels for all types of bullying except for 

relational bullying. Girls experienced high levels of relational bullying, and this was confirmed 

with the chi-square test with a p-value of 0.015 as opposed to boys. 

 

 

 

 

 

 

 

Never
44%

Seldom
13%

Sometimes
27%

Often
8%

Always
8%

Felt Dizzy



  

34 
 

Table 4.1 

Levels of Types of Bullying by Gender 

Bullying Types Boys  Girls Test P-value Different 

Cyberbullying Victims 17.02 16.34 Chi-squared 0.856 No 

Relational Bullying Victims 29.08 40.52 Chi-squared 0.015 Yes 

Verbal Bullying Victims 72.34 79.74 Chi-squared 0.089 No 

Physical Bullying Victims 40.07 43.79 Chi-squared 0.452 No 

Sexual Bullying Victims 21.63 24.84 Chi-squared 0.446 No 

 

Table 4.1 is showing the level of bullying type per gender. These two groups are mutually 

exclusive and their totals therefore do not add up to 100%.  

 

Table 4.2 below differentiates between those that had psychosomatic conditions and those that 

did not. Firstly, looking at the demographic characteristics of the study, these results show that 

across grades, the levels of psychosomatic health conditions were not statistically significantly 

different. For example: across grades, the levels of psychosomatic health conditions ranged 

from 38.3% among grade nines to 52.3% among grade elevens, but these results were not 

significantly different based on the p-value obtained from the Chi-squared test. Looking at age, 

the median and interquartile range (IQR) was similar for students with and without 

psychosomatic health conditions. As shown in the table, the median and IQR for individuals 

with psychosomatic health conditions was 16; 3. In addition, in terms of gender, results shows 

that the levels of psychosomatic health among females and males reported a statistically 

significant difference with a p-value of .009. 

 

The second part of the table looks at levels of psychosomatic health conditions across bullying 

behaviour categories. Results indicate a significantly higher level (55%) of individuals who 

had experienced psychosomatic health conditions among those who were bullied than among 

those who were not bullied. The Chi-squared results showed that the percentages were 

fundamentally different; they were statistically significantly different. For individuals who 

were and were not bullies, levels of psychosomatic conditions were relatively the same, as they 



  

35 
 

were between individuals who were and were not bully-victims, meaning that there was no 

statistically significant difference. 

Table 4.2  

Levels of Psychosomatic Health Conditions 

 

 

Tables 4.3 to 4.5 below show the relationship between the various independent variables and 

the outcome of psychosomatic health, first with bivariate logistic regression and then 

controlling for demographic variables, being age, gender, and race. The first regression table 

looks at the bivariate and multivariate relationship between psychosomatic health and being a 

bully. In terms of age, the results indicated that for every one-year increase in age, there was 

an 8% increased likelihood of having a psychosomatic health condition. The p-value was 

partially significant, as it was above .05, but below the .01 level of significance. The 

multivariate level results indicated that for every one-year increase in age, there was a 10% 

higher likelihood of having a psychosomatic health condition while controlling for bullying 

and other demographic variables. This shows that as the students’ age increased, their 

likelihood of having a psychosomatic health condition increased.  

Demographic Characteristics

Psychosomatic 

Condition Test P-value Different

Gender Chi-squared 0.009 Yes

Female 54.25

Male 41.13

Grade Chi-squared 0.346 No

Grade 8 42.14

Grade 9 38.30

Grade 10 48.09

Grade 11 52.31

Race Chi-squared 0.351 No

Black 46.54

Indian 50.00

Coloured 33.33

Age (median; IQR) 16;3 Wilcoxon 0.203 No

Bullying Behaviour

Bullied Chi-squared 0.035 Yes

No 43.07

Yes 55.21

Bully Chi-squared 0.635 No

No 45.09

Yes 47.71

Bullied-Bully Chi-squared 0.446 No

No 46.27

Yes 39.39



  

36 
 

 

In terms of race, two categories were used: black and non-black. Results show that the 

likelihood of having a psychosomatic health condition almost tripled for black learners with a 

2.6- fold higher likelihood (2.612) as compared to non-black learners at the bivariate level. 

However, at the multivariate level, black learners’ chances of having a psychosomatic health 

condition were doubled. These results for race at bivariate and multivariate levels were not 

statistically significant, meaning that race was not a predictor of psychosomatic health 

conditions. Looking at gender, when students were female, their likelihood of having a 

psychosomatic health condition was 1.697 times that of males at the bivariate level and 1.735 

times at the multivariate level. At both bivariate and multivariate levels, the p-value showed a 

statistically significant relationship. The first regression table shows that being a bully 

increased the likelihood of having a psychosomatic condition by 1% at the bivariate level and 

19% at the multivariate level. Nevertheless, whether other variables were controlled for or not, 

the association between bullying status and having a psychosomatic condition was not 

significant.  

 

Table 4.3 

Regression Results 

 

 

Table 4.4 below shows the relationship between being bullied and having a psychosomatic 

condition while controlling for demographic background factors. Again, similar results of 

association as those of Table 4.2 were found between the demographic variables and having a 

psychosomatic condition. Individuals that were bullied had a 60% higher likelihood at the 

bivariate level of having a psychosomatic health condition than those who were not bullied, 

and these results were statistically significant. In the multivariate level model, the magnitude 

of the association between individuals that were bullied and having a psychosomatic condition 

Odds Ratio P-value Odds Ratio P-value

Age 1.076 0.088 1.096 0.04

Race (Non-Black)

Black 2.612 0.101 2.051 0.241

Gender (Male)

Female 1.697 0.009 1.735 0.008

Bully (No)

Yes 1.111 0.635 1.189 0.447

BIVARIATE MULTIVARIATE



  

37 
 

decreased to a 47% higher likelihood, and the level of statistical significance decreased as well, 

to render the association no longer significant.  

 

Table 4.4 

 Relationship between being bullied and psychosomatic condition 

  

 

Table 4.5 below gives the results of the association between being a bullied-bully (bully-

victim) and having a psychosomatic condition at bivariate and multivariate levels, controlling 

for demographic variables once again. As seen in the table, age and gender had a positive and 

significant relationship with having a psychosomatic condition. Specifically, in the final model, 

for every year increase in age, the likelihood of having a psychosomatic condition increased 

by 9%, and females had a 72% higher likelihood of having a psychosomatic health condition 

than males. Additionally, Table 4.5 shows that the relationship between being a bullied-bully 

and having a psychosomatic condition was negative as well as non-significant at both bivariate 

and multivariate levels. Here, the likelihood of having a psychosomatic condition decreased 

among individuals who were bullied-bullies by 25% at the bivariate level and 28% at the 

multivariate level. The non-significance of the regression results indicates that these 

relationships were seen among the sample individuals, yet would not be able to be generalised 

to the greater population that the sample represented. 

 

 

 

 

 

 

 

Odds Ratio P-value Odds Ratio P-value

Age 1.076 0.088 1.096 0.04

Race (Non-Black)

Black 2.612 0.101 2.024 0.248

Gender (Male)

Female 1.697 0.009 1.624 0.021

Bullied (No)

Yes 1.629 0.036 1.474 0.104

BIVARIATE MULTIVARIATE



  

38 
 

Table 4.5 

 Association between being a bully-victim and psychosomatic condition 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Odds Ratio P-value Odds Ratio P-value

Age 1.076 0.088 1.091 0.048

Race (Non-Black)

Black 2.612 0.101 2.004 0.255

Gender (Male)

Female 1.697 0.009 1.723 0.009

Bullied-Bully (No)

Yes 0.755 0.447 0.724 0.391

BIVARIATE MULTIVARIATE



  

39 
 

CHAPTER 5: DISCUSSION 

 

5.1 Levels of Psychosomatic Conditions 

 

The present study found that 46% of learners from Gauteng township (Orlando West) high 

schools suffered from psychosomatic health conditions, while 54% of the sampled population 

experienced no psychosomatic health conditions. Psychosomatic health conditions in this study 

included difficulty concentrating, difficulty sleeping, headache, stomach ache, feeling tense, 

loss of appetite, feeling sad, and feeling dizzy.  

 

The results for those learners who reported that they always or frequently suffered from the 

various specific psychosomatic conditions were as follows: 

• Difficulty in concentrating: 6% often; 3% always. 

•  Difficulty in sleeping; 5% often; 5% always. 

• Headaches: 8% often; 7% always. 

• Stomach ache: 9% often; 3% always. 

• Feeling tense: 5% often; 4% always. 

• Loss of appetite: 11% often; 5% always. 

• Feeling sad: 5% often; 5% always. 

• Dizziness: 9% often; 3% always. 

 

Previous studies showing levels of psychosomatic conditions among adolescents have shown 

similar results to those found in this study. A study conducted in urban and rural middle schools 

of China found that children who suffered from bullying often or always had a headache, 16.1% 

that they often or always experienced sleep problems, and 16.5% that they often or always had 

abdominal pains (Li, Sidibe, Shen, & Hesketh,, 2019). This study seems to have found slightly 

higher levels of results as compared to the current study; this might be due to the high number 

of participants in the Chinese study as compared to the current study, as well as the as the fact 

that the current study focused on only one township area in Soweto. 

 

According to a study conducted in Portugal, looking at the symptoms of the studied group, it 

was found that there was no statistically significant difference between each symptom of 

psychosomatic health surveyed: stomach ache (X(3) 2 = 16.70); p = .16 (2-sided); nervousness 



  

40 
 

(X(3) 2 = 12.08); headache (X(3) 2 = 20.80); irritability or bad temper (X(3) 2 = 8.94); p = .71 

(2-sided); backache (X(3) 2 = 5.47); p = .22 (2-sided); feeling low (X(3) 2 = 20.19); p = .06 

(2-sided); and depression (X(3) 2 = 14.21); p = .29 (2-sided); (Seixas, Coelho, & Nicholas-

Fischer, 2013). 

 

A study conducted in  Iran reports a strong association between bullying behaviour and 

psychosomatic conditions. In this case, sadness, feeling low, anxiety, and sleep difficulties 

were the most prevalent psychosomatic problems experienced by students involved in bullying 

(Rezapour, Soori, Nezam Tabar, & Khanjani, 2020). This study supports the findings of the 

current study. However, it had a large sample of 834 participants, which is well over double 

the number of participants in this study. This indicates that the association between bullying 

and psychosomatic health is strong and affects most school-age children even though most 

cases are not reported, which is a concern as the well-being and academic performance of the 

students are affected. Future studies should therefore include a greater sample size in order to 

provide a better distribution and to be able to investigate detailed gender and age effects. 

 

5.2 Levels and Gender Differentials of Bullying 

 

For the second objective of the current study (levels of types of bullying by gender), this study 

found that bullying types were generally on the same levels for both genders, except for 

relational bullying. This correlates with previous studies that have also shown that relational 

bullying is higher among girls, such as the United States study which otherwise reported that 

there were no significant differences based on the z tests in the correlation of bullying across 

gender (Carbone-Lopez, Esbensen, & Brick, 2010). Additionally, according to Menesini and 

Salmivalli (2017), girls are more likely to be involved in relational or verbal bullying as 

compared to boys. 

 

The reasons given in previous studies for why relational bullying is high among girls are that 

most girl bullies engage in relational bullying due to fear of being targets or actually being 

bullied themselves, or relationship problems with disloyal friends (Besag, 2006; Owens et al., 

2000). Additionally, according to Underwood and Rosen (2011), it is sometimes because girls 

often lack conflict management skills and also because this form of bullying (relational 

bullying) is socially acceptable for girls. Other reasons include the creation of excitement in 

their lives, competition, jealousy, and boredom alleviation (Moore et al., 2017). 



  

41 
 

 

The current study supports the study results by Sharrif (2008), who reported that boys and girls 

display similar levels of bullying, although boys are more exposed to open attacks and girls to 

verbal and relational bullying. According to Shelley and Peterson (2019), males more than 

females on average are likely perpetrators and victims of bullying focusing on gender 

differences; however, there are different types of bullying that boys and girls are more involved 

in, directly or indirectly. Females are more involved in indirect forms of bullying (mostly 

inflicting harm by damaging another’s social reputation, peer relationships, and self-esteem) 

while boys are involved in direct forms of bullying (for example passing on insults or spreading 

rumours. (Besag, 2006). According to Tokunaga (2010), the cyberbullying research results are 

vague, with some research studies suggesting similarity in the genders and others suggesting 

greater prevalence among females (Cross et al., 2014; Viljoen, O’Neil, & Sidhu, 2005).  

 

A study conducted in Brazil schools found a significantly greater difference in the frequency 

of involvement in bullying between girls and boys. These results show that boys were more 

frequently involved in bullying behaviours, either as bullies, bully-victims, or victims (Isolan, 

Salum, Osowski, Zottis, & Manfro, 2013). This result may differ from the current study due to 

its low number of female participants in the study as compared to males. Future studies could 

improve on any shortcomings if more girls were encouraged to take part in the study and 

included in the sample to ensure approximately equal numbers of males and females. 

 

Research by Mishra, Thapa, Marahatta, and Mahotra (2018) reports that bullying was exhibited 

less by girls than boys. Among girls, there were relatively fewer bullies and bully-victims as 

compared to victims of bullying. On the other hand, the percentage of boys who were bullied 

was 62.76%, similar to the 62.25% of boys that bullied others. 

 

This contrasted with the results of another study conducted in Ireland which found that the 

prevalence of victims among both sexes was similar at 25.1%, while the prevalence of bullies 

at school in the previous couple of months had been 13.3%.  

 

Regarding regression results for gender, this study revealed that female students had a 

significantly higher likelihood of 62–74% of having a psychosomatic condition. A recent study 

conducted by Shelley and Peterson (2019) in the United States also indicated that females who 

were bullied showed a higher level of psychosomatic symptoms than males.  



  

42 
 

 

5.3 Association between Bullying and Psychosomatic Conditions 

 

The main aim of this study was to establish the association between various types of bullying 

behaviour and psychosomatic conditions. According to the final multivariate logistic 

regression model, this study showed that bullies had a 19% higher likelihood of having a 

psychosomatic condition, bully victims had a 47% higher likelihood, while bullied-bullies had 

a 28% lower likelihood of having a psychosomatic condition. Our results, though insignificant 

of the size, give insights on the relationship between bullying behaviour and psychosomatic 

conditions. As can be seen, whether individuals were bullies or bullied (victims of bullying), 

the association with having a psychosomatic condition was positive. Also, the magnitude of 

the association was greater among victims than perpetrators. This indicates the detrimental 

nature of bullying behaviour on health and supports the need for prevention campaigns and 

advocacy against bullying behaviour in high schools. The positive association between 

bullying victims and perpetrators found in this study aligns with results from previous studies 

in different contexts as shown below. 

 

This study was consistent with findings by Gini (2008) that children who were being bullied 

had a significantly higher risk of psychosomatic problems. However, our study reported that if 

one is a bully-victim, there is no significant relationship with having psychosomatic health 

conditions. These results are contrary to the same study by Gini (2008), which reported a 

significantly higher risk for psychosomatic problems for bully-victims. Additionally, according 

to a study conducted by Singh and Edbor (2019), bullied adolescents are more likely to 

experience social withdrawal, poor sleep, anxiety, academic backwardness, depression, 

aggression, and loneliness. 

 

A study conducted in Netherlands found that there is a significantly higher chance for bully 

victims to experience psychosomatic health symptoms (such as abdominal pain, depression, 

headache, and feeling tired) in consequence as compared to children not involved in acts of 

bullying (Fekkes, Pijpers, & Verloove-Vanhorick, 2004). Italian research using meta-analysis 

also supports the results of this study and confirms that a higher risk for psychosomatic 

problems was found in children who were bullies. Victims and bully-victims show the largest 

effect; however, although bullies were at lower risk for psychosomatic health problems, they 

were still were predisposed (Gini & Pozzoli, 2009, 2013). Another study in Nepal found that 



  

43 
 

both bullies and victims of bullying tended to experience physical as well as psychological 

symptoms, including psychosomatic symptoms which could continue into adulthood (Mishra, 

Thapa, Marahatta, & Mahotra, 2018). Furthermore, Li, Sidibe, Shen, & Hesketh, (2019) found 

that the prevalence of victims who were also bullying was significant among adolescents and 

was linked to psychosomatic conditions. 

 

According to Sourander and Ikonen (2010), the relationship between traditional bullying 

victimisation and psychosomatic, physical health, and psychological problems has been shown 

in a few longitudinal studies and a number of cross-sectional studies. The studies mentioned 

above show that individuals that experience bullying behaviour tend to have a higher likelihood 

of psychosomatic conditions. According to Gini and Pozzoli (2009), the possible reasons that 

victims and bully-victims are at higher risk for a variety of psychosomatic conditions is that 

they have similar difficulties in numerous spheres such as health problems, low emotional 

adjustment, and poor relationships with their classmates. In addition, there are similarities and 

differences between bullies and victims. Both bullies and victims are characterised by poor 

academic performance. In contrast, bullies display externalised problems such as alcohol and 

drug abuse, as well as poor school adjustment. On the other hand, victims often report 

experiencing anxiety, depression, loneliness and low self-esteem. In addition, bully victims are 

labelled as being hyperactive, some being anxious, and/or suffering from poor social 

adjustment. 

 

According to Fekkes, Pijpers and Verloove-Vanhorick (2004), numerous studies have reported 

that parental use of physical punishment and harsh discipline is definitely associated with 

children exhibiting more bullying behaviour towards their peers. Additionally, the stress in an 

individual due to being bullied is considered to contribute to the development of psychosomatic 

conditions. Being bullied also contributes to a higher likelihood of depression and health 

symptoms.  

 

The findings of this study are consistent with other studies around the world that have 

demonstrated poor social adjustment, high levels of social isolation, anxiety, health problems, 

attention deficit hyperactivity disorder and disturbed personalities in bully-victims 

(Kumpulainen, Räsänen and Puura, 2001; Gemzøe and Einarsen, 2002.). For instance in 

Sweden, victims in general were found to be very problematic as they exhibited a variety of 

health problems and difficulties in the interpersonal domain, such as high levels of social 



  

44 
 

anxiety, loneliness, fear of negative evaluation, avoidance of social situations and social skills 

deficits (Kumpulainen et al., , 2001). 

 

The study intended to test the theory mentioned in Section 2.11.1. According to that theory, 

bullies are likely to display negative and antisocial behaviour, for example, truancy, 

delinquency, and substance abuse during adolescence, and are at risk for psychiatric disorders. 

Furthermore, bully-victims were likely to suffer from a variety of psychosomatic and 

behavioural problems, for example tiredness, nervousness, sleeping problems and dizziness. 

According to the Social Dominance Theory, bullying was more common among boys than girls 

while psychosomatic health challenges happened more among girls than boys. According to 

the STD, bullying is usually used, especially among boys, to increase social dominance status, 

and once this is gained bullying behaviour declines. This study’s findings are consistent with 

what the theory argues. Therefore, one can conclude that the STD is applicable in the South 

African context. 

 

5.4 Limitations and Future Directions 

 

The current research study provides information about the relationship between bullying 

behaviour and psychosomatic conditions among township learners. It should be noted that this 

study has a few limitations which could be addressed in future research. Firstly, it focused on 

only three township schools, in Orlando West, Soweto. It must be noted that these results do 

not include other schools from other townships and therefore are not an indication of the 

situation in all townships in Gauteng. To overcome this limitation, further research could entail 

a larger study that incorporates more schools across different townships in Gauteng to get a 

more holistic picture of the association between the independent and dependent variables. 

 

Secondly, this study was quantitative, which might have limited some children’s ability to 

express their emotions, as well as prevented the researcher from understanding the reasons 

underlying their answers. Additionally, the study methodology may have limited children from 

expressing their experience of bullying as respondents had to answer the questionnaire 

according to scales. Perhaps a qualitative or mixed-methods approach could be helpful to 

understand the reasons behind the association between bullying and psychosomatic conditions. 

 



  

45 
 

Thirdly, the questionnaire for the current study was very long to answer and due to children’s 

short attention span, extra time was needed to calm participants down persistently for them to 

focus on filling out and completing the research questionnaire. As a result, many did not 

complete the questionnaire and had to rush off to other subject periods. In future, this should 

be considered when creating a questionnaire specifically for young people, and the time needed 

should also be communicated to the teachers that are allocated to work with researchers to 

alleviate these problems. 

 

Lastly, when conducting the questionnaire, most children were sitting in pairs, which might 

have caused them to influence each other’s results. Some might have been uncomfortable 

disclosing their experiences of bullying, as they might have been afraid of how their peers 

would perceive them. Additionally, they may have been scared that their bullies would discover 

their truthful resp