Professional Documents
Culture Documents
Rabia Khawar
006-GCU-PHD-PSY-09
2009-2012
Department of Psychology
GC University Lahore, Pakistan
Effect of Mood on Emotion Recognition and Quality of Friendship among
By
Rabia Khawar
006-GCU-PHD-PSY-09
2009-2012
Department of Psychology
GC University Lahore, Pakistan
Dedicated to
My Beloved Family
CONTENTS
Topics Page No.
List of Tables ………………………………………………………………………..........i
List of Figures ……………………………………………………………………………vii
List of Appendices ………………………………………………………………….........ix
Acknowledgements ………………………………………………………………………x
Abstract …………………………………………………………………………………..xiii
Results………………………………………………………………....102
Discussion……………………………………………………………..131
Section B. Friendship Qualities Scale…………………………………………………139
Results………………………………………………………………....147
Discussion……………………………………………………………. 171
Section A: Study 3
Phase 1: Development of Mood Induction Procedures (MIPS)...................................217
Objectives of the Study………………………….................………….218
Hypotheses……………………………………..……………………...218
Method……………………………..………………………………….218
Result…………………………………………………………………..227
Discussion……………………………………………...……………...262
i
Table 2.17 Descriptive Statistics for Bullying and Victimization Scales 122
Table 2.18 MANOVA showing Gender and Grade differences for total bullying 123
and victimization
Table 2.19 Post Hoc Comparisons among Grade Levels on Victimization and 126
Bullying
Table 2.20 Bullying Role Classification Following OBVQ and Teacher 128
Nomination Criteria
Table 2.21 Multinomial Logistic Regression Predicting Accuracy of Self- 129
reported Bullying and Victimization
Table 2.22 Demographic Characteristics of Samples for FQS. 147
Table 2.23 Inter-Item Correlation between English and the Urdu Versions of 149
FQS.
Table 2.24 Reliability Coefficients for English and Urdu Versions of FQS 150
(Sample I; N = 32)
Table 2.25 Kaiser-Meyer-Olkin Test of Sampling Adequacy and Bartlett Test 151
of Sphericity for FQS
Table 2.26 Principal Component Analysis with Varimax Rotation for FQS Urdu 152
Version (N = 151).
Table 2.27 Shapiro-Wik Test for Normality 154
Table 2.28 Factor Matrix for FQS Urdu using Principal Axis Factoring (N = 155
151)
Table 2.29 Pattern Matrix of Principal Axis Factoring with Promax Rotation for 157
FQS Urdu Version
Table 2.30 Descriptive and Item-Total Statistics of the FQS Urdu Version (N = 159
151).
Table 2.31 Reliability Analysis of the Friendship Qualities Scale (FQS) Urdu 160
Version and Subscales Derived from EFA (N = 151).
Table 2.32 Inter-correlations and Convergent/Discriminant Validity of the 4 161
Factor FQS Urdu Version (N = 135)
Table 2.33 Missing Value Analysis for Items on FQS 163
Table 2.34 Descriptive Statistics for Items on FQS 164
ii
Table 2.35 Fit Indices of FQS Models 167
Table 2.36 Standardized Factor Loadings on CFA Model 2 for FQS 168
Table 2.37 Means, Standard Deviations and Alpha Coefficients for FQS 169
Subscales
Table 2.38 Inter-correlations among FQS Factors and the Total Score 170
Table 3.1 Distribution of Bully-Victim Groups according to the Type of 184
Schools
Table 3.2 Pearson Chi-Square for Bully-Victim Groups across Gender, Grades 185
and type of Schools
Table 3.3 Severity of Bullying Behavior by Bully-Victim Groups, Gender and 186
Grades
Table 3.4 Severity of Victimization Experience by Bully-Victim Groups, 187
Gender and Grades
Table 3.5 Categorical Distribution of Victimization Types across Gender and 188
Grade Level
Table 3.6 Pearson Chi-square for the Types of Victimization across Gender 189
and Grades
Table 3.7 Class Reported by Victims and Bully-Victims in which the Bullies 191
Study
Table 3.8 Number of Bullies Reported by Victims and Bully-Victims across 192
Gender and Grades
Table 3.9 Pearson Chi-square of Bully’s Class and Number across Gender and 193
Grades
Table 3.10 Duration of Victimization Reported by the Total Sample and Those 193
who had been Bullied (Victims and Bully-victims)
Table 3.11 Place of Bullying reported by Gender and Grade 195
Table 3.12 Frequency Distribution of the Boys and Girls and Students of 196
Different Grades about Disclosure of Bullying
Table 3.13 Proportion of Participants Bullying Others by Types of Bullying 198
Table 3.14 Summary of t-test for Boys and Girls on OBVQ Items (N = 817) 200
Table 3.15 One-Way ANOVA Showing Grade Differences on OBVQ Items (N =817) 201
iii
Table 3.16 One-Way ANOVA Showing Mean Differences on OBVQ Items (N 202
=817) across Bully-Victim Groups
Table 3.17 Means, Standard Deviation, Skewness and Kurtosis for Academic 203
and Demographic Variables for Total Sample
Table 3.18 ANOVA for Academic Variables across Gender and Bully/Victim 204
Groups
Table 3.19 Post Hoc Scheffe Test for Multiple Comparisons across 206
Bully/victim Groups for Academic Achievement
Table 3.20 Correlation of Socioeconomic Variables with Bullying and 208
Victimization
Table 3.21 Frequency, Percentages, Skewness and Kurtosis for Parental Level 209
of Education
Table 3.22 MANOVA for effect of Parental Education on Involvement in 210
Bullying and Victimization
Table 4.1a Descriptive Statistics for Ratings of Video Songs by Children (n = 228
15)
Table 4.2a Descriptive Statistics for Ratings of Instrumental Music by Children 229
(n = 15)
Table 4.3a Descriptive Statistics for Ratings of Sad Pictures by Judges (n = 10) 231
Table 4.4a Descriptive Statistics for Ratings of Happy Pictures by Judges (n = 232
10)
Table 4.5a Descriptive Statistics for Ratings of Neutral Pictures by Judges (n = 233
10)
Table 4.6a Characteristics of MIP Groups (N = 48). 235
Table 4.7a Descriptive Statistics for Mood Induction Trials within Positive MI 236
Group (n = 32)
Table 4.8a Descriptive Statistics for Mood Induction Trials within Negative MI 237
Group (n = 32)
Table 4.9a MANOVA for Valence of Techniques across MI Groups and 238
Gender
Table 4.10a MANOVA for Level of Arousal of Techniques across MI Groups 241
iv
and Gender
Table 4.11a Mixed Factorial ANOVA for Mood Valence in Positive MI Group 245
Table 4.12a Paired Comparisons of Valence Scores for Happy MI Techniques 248
Table 4.13a Mixed Factorial ANOVA for Mood Arousal in Positive MI 249
Condition
Table 4.14a Paired Comparisons of Arousal Scores for Happy MI Techniques 253
Table 4.15a Mixed Factorial ANOVA for Evaluating Change in Valence in 254
Negative MI Condition
Table 4.16a Paired Comparisons of Valence Scores for Sad MI Techniques 257
Table 4.17a Mixed Factorial ANOVA for Evaluating Change in Arousal in 258
Negative MI Condition
Table 4.18a Paired Comparisons of Arousal Scores for Sad MI Techniques 261
Table 4.19a Gender Distribution across Bully/Victim Groups 277
Table 4.20a Distribution of Bully-Victims Groups in Four Mood Conditions 278
Table 4.21a Descriptive Statistics of FEEL Test and Individual Emotions 279
Table 4.22a Normality of Responses on Individual Emotions and FEEL Total 280
Score
Table 4.23a Between-Within ANOVA showing Effects of Mood Induction on 281
Subjective Mood Ratings across Conditions
Table 4.24a ANOVA showing Effects of Involvement in Bullying Behaviors and 284
Gender on Overall Emotion Recognition Ability
Table 4.25a Post Hoc Comparisons among Bully/Victim Groups for FEEL Test 287
Table 4.26a Between-Within ANOVA for Emotion Categories on FEEL for 289
Bully/Victim Groups, Mood Condition and Gender
Table 4.27a MANOVA showing Effect of Mood on Emotion Recognition 291
among Bully-Victim Groups
Table 4.28a Post Hoc Comparison among Bully/Victim Groups for Individual 293
Emotions on FEEL
v
Table 4.30a Post Hoc Comparisons among Bully/Victim Groups for Reaction 298
Time on FEEL
Table 4.1b Distribution of Bully/Victim Groups across Gender (N = 672) 306
Table 4.2b Univariate Analysis of Variance of Friendship Quality for 307
Bully/Victim Groups, Gender and Grade
Table 4.3b Post Hoc Comparisons of Bully/Victim Groups for Quality of 308
Friendship
Table 4.4b MANOVA showing Main Effects and Interaction of Bully/Victim 312
Groups and Gender on FQS Subscales and Conflict
Table 4.1c Academic and Socioeconomic Correlates of Friendship Quality and 318
Emotion Recognition
Table 4.2c Correlation between Friendship Quality and Emotion Recognition 319
across Bully/Victim Groups
Table 4.3c Regression Analysis for Emotion Recognition Ability and 320
Friendship Quality on Global Victimization (N = 370)
Table 4.4c Regression Analysis for Emotion Recognition Ability and 321
Friendship Quality on Global Bullying (N = 370)
vi
LIST OF FIGURES Page
Figure No. No.
vii
Figure 4.10a Sad and Neutral Valence Scores across Mood Techniques 256
Figure 4.11a Level of Arousal for Sad and Happy Mood Conditions across MI 259
Techniques
Figure 4.12a Mean Arousal Scores of Four MI Techniques across Blocks within 260
Negative MI Group
Figure 4.13a The Procedure of Random Assignment 273
Figure 4.14a The Experimental Process 275
Figure 4.15a FEEL Test Procedure 276
Figure 4.16a Pre and Post Induction Mood Ratings across three Mood 283
Conditions
Figure 4.17a Mean Scores of Bully/victim Groups on FEEL Test 285
Figure 4.18a Mean Scores of Boys and Girls on FEEL Test across Bully/victim 286
Groups
Figure 4.19a Mean differences of Bullly/Victim Groups on Anger across Mood 295
Conditions
Figure 4.20a Mean differences of Bullly/Victim Groups on FEEL across Mood 296
Conditions
Figure 4.21a Mean Reaction Time on FEEL across Bully/Victim Groups 299
Figure 4.1b Estimated Marginal Means of Friendship Quality across 309
Bully/Victim Groups
Figure 4.2b Mean Scores of Boys and Girls on FQS 310
Figure 4.3b Gender and Grade Interaction for Friendship Quality 311
Figure 4.4b Estimated Marginal Means of Conflict across Gender among 314
Bully/Victim
viii
LIST OF APPENDICES
Contents Page No
App. Label
ix
Acknowledgements
Finally, 2015 marks the end of a long process and this would not have been possible
without help from some very important people who have facilitated and supported me
First, and foremost, I would like to acknowledge my supervisor Prof. Dr. Farah
Malik. She has the genius to explore the novel avenues in the field of psychology in
through the difficult times of conceiving the project and collecting the data. Her skillful
guidance continued in conducting major analysis. I will be forever grateful to her for her
years of mentoring that evoked the insight and passion for research in me.
Shahida Batool, for her valuable feedback, and constructive advice that provided me the
right direction to make this work to come to a good end. Her suggestions have refined
my communication skills, and her professional discipline and diligence inspire me to the
same. She has been very positive and generously gave of her valuable time and vast
knowledge.
cooperation, patient listening to the problems and unconventional support when I needed
it the most.
financial assistance that kept me going smoothly during the course of research. I also
x
Chancellor, the Registrar, the Dean of Social Sciences and Incharge Department of
Hoffmann, CSEAMedia) for providing and allowing me to use and translate the
their schools. I am also obliged to all the students and their parents for sharing the
information. My special thanks are extended to The Punjab School Township, Lahore for
taking keen interest in the research and arranging special awareness campaign about
Department, GCU Lahore especially Mr. Maroof, and all the others who always
facilitated me.
I am also indebted to my close friends Rabia Iftikhar and Nighat Zeeshan for all
the good times and critical reflections, love, support, and companionship especially
Musaffa Butt for his invaluable suggestions and help. These friends have always been a
I thank Allah Almighty for blessing me with the parents who have shown their
unconditional love and provision throughout my life. They have made substantial
appreciation of them. Thanks to my brother also for all the sharing and love. I can never
xi
forget my first teacher and mentor, my grandfather (late) who left me during my research.
He always believed in me and my ambitions that kept my spirits high. God bless his soul.
I would not have accomplished anything without continual and generous patience,
love, care, and support of my beloved husband, Dr. Zulfiqar Ali. I am blessed to have
you in my life. Thanks to my lovely kids Hassan and Hareem who deserve much of the
credit in helping me. Their laughter always eased my pains and relieved my worries and
Rabia Khawar
xii
ABSTRACT
Information Processing and Social Cognitive dimensions of bullying (Crick & Dodge,
1994; Lemerise & Arsenio, 2000; Sutton, 2000). Following the above mentioned
connotations, the current research endeavor mainly focused on effect of mood on emotion
recognition ability of school bullies and victims for its implications in SIP Model.
Besides estimating the occurrence of bullying in 4th, 5th and 6th graders, friendship quality
of bully/victim groups was also evaluated in view of the role played by peer relationships
and Friendship Qualities Scale (FQS; Bukowski, Hoza & Boivin, 1994) were translated
into Urdu and were evaluated for the construct validity. Consistent with the original
model, both Exploratory and Confirmatory Factor Analysis (EFA and CFA) afforded a
two factor (Bullying and Victimization) solution for OBVQ. Whereas, contrary to the
original 5 factor model of friendship quality, a four factor model (approval, intimacy,
help and conflict) was confirmed for FQS. The second study provided an account of
prevalence of bullying within a sample of 817 students (46.1% boys & 53.9% girls; Mage
= 10.89, SDage = 1.12). Only 34.6% students were found uninvolved in any kind of
bully/victim groups. As a prerequisite for the study, we first experimentally evaluated the
xiii
combination were found more effective than other techniques and therefore considered
suitable for the main study. During the second phase, participants (N = 394; Mage = 10.77;
SDage = 1.07) were randomly assigned and primed with sad, happy and neutral mood.
Emotion Labeling (FEEL) test of emotion recognition. Contrary to assumptions mood did
condition scored higher on FEEL test and recognized the anger more accurately than
those in sad or happy mood condition. Moreover students involved in various bullying
(N = 672; Mage = 10.86; SDage = 1.08). ANOVA results showed that bully-victim group
had significantly poorer quality of friendship than other groups. Specifically, they
showed poorer conflict resolution and lesser intimacy as compared with uninvolved
significantly poorer in resolving conflict as compared with boys. With regard to academic
Lastly we found a significant yet weaker relationship between friendship quality and
emotion recognition ability of students (r = .11, p < .05). Both of these variables proved
xiv
1
Chapter-I
INTRODUCTION
and well-being of the students across different countries (Chibbaro, 2007; Thompson &
habitual and repeated behavior, meant to impose domination over the weaker individual.
Bullying behavior primarily revolves around coercion and intimidation with either direct
aggression such as slapping, name calling, pushing, swearing, and damaging among
disparate others or using indirect forms, for example gossiping, rumors and social
rejection which exercises immense influence and causes harm to the victims (Beran &
Lupart, 2009).
Bullying associated to different forms of violence and aggression can lead to high
profile incidents of escalated violence later in life (Ttofi, Farrington, & Losel, 2012). So
dealing with this issue at earlier stage can not only create safer and healthier school
environment but would also affect the community at large. At individual level, bullying
may result in physical, academic and psychological harm to those who are involved in it
Researchers have highlighted the impact of bullying upon children’s cognitive, affective
and social development and competence (Arsenio & Lemerise, 2002; Wal, Wit, &
Hirasing, 2003; Wong, Lok, Lo, & Ma, 2008).The present study broadly intended to
identify nature and extend of bullying and its emotional and social repercussions among
A considerable change has been observed in the term bullying over time. The first
‘The Times’ in 1862. Subsequently, Burk (1897) wrote the first noteworthy research
paper about teasing and bullying in youth. By the half of 20th century, researchers
seriously taken up as a research matter in Scandinavia in mid1970s after Pikas (1975) and
Olweus (1978) made first systematic attempt to empirically investigate the problem. It
turned out to be a societal and official concern after a newspaper in 1982 reported suicide
of three Norwegian boys as a result of school bullying (Olweus 1993). Olweus defined
bullying as an aggressive act comprising both direct and indirect strategies deliberately
repeated over time by more powerful individuals to hurt weaker ones. Since then bullying
has become the focus of many investigators who had developed variety of approaches
aggression and bullying but also individual and group bullying/mobbing (Olweus, 1993),
a distinction that has been later overlooked. Besag (1989) emphasized the bullies'
among those who are powerless to resist, for some specific gain or gratification. Espelage
and Asidao (2001) discussed three dimensions that differentiate bullying from
picking on weaker and controllable targets to hurt them many times over by
The most widely accepted definition of bullying had been offered by Olweus
and stronger than the victim. Olweus (1994, 1999) distinguished bullying from ordinary
fight or quarrelling between two parties of equal strength by emphasizing the notion of
power imbalance. It’s a combination of certain direct and indirect negative actions
performed by one or more students repeatedly over time with an intention to harm weaker
students. Victims’ inability to defend him or her is another important feature of Olweus
definition.
more powerful ones who intentionally inflict harm to them through physical, verbal and
psychological means.
Heald (1994) reported the British legal definition of bullying that describes it as
person or put him under stress by conducting physical or psychological violence over an
extended time period. It brings feelings of insecurity to the victim as target is usually
standing violent acts from low to high intensity that are planned by a stronger person to
hurt the vulnerable target. Rigby’s (2002) extensive discussion on defining bullying
concludes that it involves an urge to hurt followed by hurtful action which is taken in
cases where power imbalance exists between parties and the more powerful person
repeatedly uses his/her power in an unjust manner. Moreover the perpetrator manifests
enjoyment and pleasure while the target experiences a sense of being oppressed. Hence
two components are critical while defining bullying; a) the feeling on part of bully and b)
Most recently Division of Violence Prevention, Centers for Disease Control and
Prevention (CDC; Gladden, Kantor, Hamburger & Lumpkin, 2014, pp 7), released a
youth who are not siblings or current dating partners that involves an observed or
repeated. Bullying may inflict harm or distress on the targeted youth including physical,
researchers (Swearer, Siebecker, Johnsen- Frerichs, & Wang, 2010). Yet, these
(c) Long-standing and repetitive acts of direct and indirect aggression and
Bullying is complex social process and those who are involved in it play different
roles (Cohen, Garcia, Apfel & Master, 2006). Salmivalli, Lagerspetz, Björkvist,
Österman, and Kaukiainen (1996) had discussed six different participant roles (bullies,
victims, bully-victim, bully reinforces, bully assistants, defenders of victim, and the
distinguish among perpetrators as bullies, targets as victims, bully-victim and those who
are not identified in any of these three roles as not involved (Olweus, 1995; Solberg &
Olweus, 2003). Further variations have been rarely addressed in bullying investigation.
Bullies
Bullies, who frequently and diligently hurt others over a period of time, are often
recognized for exhibiting hostile yearnings and physical strength. They are considered
compassionate (Olweus, 1995). Olweus (1997) later described them as less anxious,
fairly secure and having normal levels of self-esteem. Additionally, children’s bullying
behavior is mostly found steady across contexts such as home and school (Lane, 1989).
Bullies are also more likely to exhibit oppositional propensities towards adults, described
as oppositional and rebellious toward adults, disruptive and antisocial, and often involved
in breaking school rules (Batsche & Knoff, 1994; Olweus, 1993, 1997).
6
Students who bully others mostly do so to get noticed by their peers and enjoy the
sense of powerful status amongst them (Sullivan, 2000). They possess positive approach
towards violence and their bullying behavior is strengthened when it is not challenged by
peers or reported to the authorities. Bullies are usually unpopular but, unlike victims, may
more systematic and clever, thus maintain a popular status among peers and supposed to
perform well in academics and tend to be socially skilled and smart to influence others
according to their own will (Caravita , DiBlasio, & Salmivalli, 2009; Dijkstra, Wilgen,
Versteegen & Fleuren, 2009; Juvonen, 2014, Veenstra, Lindenberg, Munniksm, &
Dijkstra, 2010).Because they are equally popular among teachers and students, therefore
are often difficult to identify due to their tactful manipulation. Generally, they lack
Victims
These students as the targets experience frequent bullying, for example two to
three times a month or more and are categorized as victims. They have been found to be
more anxious, alone, depressed, timid and apprehensive and physically weaker than
bullies. They tend o be less assertive, have low self-esteem, and lack confidence. These
attributes lead to poor social skills and consequently they become unable to react and
defend themselves (Olweus, 1995; 1997). They often exhibit more internalizing problems
compared to students in other bullying roles (Reijntes, et al., 2010). Low socioeconomic
status, distinct ethnic group, physical disabilities, learning difficulties and school shifting
The targets that are usually recognized as passive or pure victim never initiate the
bullying act and remain unable to retaliate, while there is another group known as
Bully/Victims
Olweus (1978) first identified this group named as provocative victims who make
unsuccessful efforts to fight back, when they are bullied. There are few who succeed in
their retaliation and have been referred to as bully-victims subsequently (Boulton &
Smith, 1994).Another term reactive bullying has emerged as a subtle form referring to the
bullies who initiate the discord and get attacked as a result of that strife. Such a bully is
an instigator who persistently teases and offends some students until that student is
provoked to respond aggressively. The reactive bully then blames the student for the
conflict (Smokowski & Kopasz, 2005). Bully-victim group may also comprise these
reactive bullies.
purposeful than those of a provocative victim (Schwartz, Proctor, & Chien, 2001).
Therefore nature and extent of aggression should be taken into account while addressing
the bully/victim group. This third group actually acts as both perpetrator and target of
bullying. They exhibit a combination of both anxious and aggressive behavior patterns,
and often being hyperactive and more self-destructive as compared to bullies or victims
(as cited in Olafsen &Viemerö, 2000). They are unpopular due to their irritability,
Uninvolved or Bystanders
Students who are not directly involved into active roles mentioned above had
1. Those who do not initiate the activity but actively take part in bullying. They
join the perpetrators in bullying the targets. They are named as followers or
assistants of bullies.
2. There is another group which does not participate in bullying behavior, yet act
as they may witness the act silently. They could also support the bullies by
3. Defenders are the students who not only dislike bullying but also try to protect
or help the target by providing direct assistance to the victim by intruding the
social behavior that is positively associated with high levels of empathy and
& Salmivalli, 2010; Gini, Albiero, Benelli, & Altoè, 2007; Juvonen, 2014).
Still there are few students who remain passive or completely irrelevant of the
scenario and considered uninvolved. They are called outsiders neither bully nor being
Bullying can take several forms such as physical, verbal, social or relational
Physical
The term demonstrates the use of physical power by the offender against the
punching, kicking, beating up, tripping, pushing and shoving; (b) damaging the
belongings of the victim for example snatching, stealing and wasting or destroying the
books, lunch, money or other goods; (c) pressurizing the target to perform an unwanted
Verbal
and drawings (Martinez-Criado, 2014). This is also the most common type of bullying,
yet goes unidentified owing to the ease and rapidity with which it occurs (Smokowski &
Kopasz, 2005). Verbal bullying often proceeds to other forms of bullying (Liepe-
Social/Relational
This type of bullying has been defined by behaviors that are intended to harm the
target’s social relationships and feelings by deliberately ignoring him or her or excluding
from the peer group (Crick & Grotpeter, 1995). This type also involves gossip, and false
rumor spreading about the victim that causes embarrassment and destroys the social
image and standing of the target (CDC, 2012; Schäfer, Werner, & Crick2002; Wolke,
Cyberbullying
Cyber bullying has caught the attention of researchers in recent years. This
particular type of bullying is mediated by digital technology perpetrators use overt and
covert ways to harm their target (Kiriakidis & Kvoura, 2010). With the help of cell
phones, emails, chat rooms and social networking, youth is excessively getting involved
in physical, verbal and relational bullying. It can happen anywhere and any times and
excluding others online, spreading rumors, distorting others image either through direct
remarks or using fake IDs (Hinduja & Patchin, 2010; Ybarra & Mitchell, 2004). It
endangers the social reputation of the victim and most adverse aspect is that the bully can
These types can be broadly grouped together into two distinct modes of direct and
indirect bullying. Direct bullying acts are blunt, explicit, overt and straightforward, often
11
demonstrated in the presence of the victim (Peskin, Tortolero, & Markham, 2006).
Indirect bullying is more covert and therefore mostly includes relational and cyber forms
of bullying (CDC, 2014). Direct bullying is often observable and includes physical
aggression (hitting, kicking), verbal aggression (teasing, racial or sexual comments), and
perpetrator’s attempts to isolate the target by forcefully keeping him or her away from
social activities. Indirect bullying can also be physical (instigating someone to attack
someone else), verbal (false gossiping), and nonverbal (excluding from a group or
activity) (Boulton & Underwood, 1992; Olweus, 2001; O’Moore & Kirkham, 2001;
Raskauskas & Stoltz, 2007). Cyber bullying in direct mode may include damage to
forming bullying groups to directly assault someone. In its indirect mode it incorporates
attempts by the bully to socially isolate the victim using relational means in a systematic
manner.
synchronized and the systematic bullying prevention program (Batsche, 1997; Ross,
1996). Time constraints, manpower and expenditure are some of the important features to
overall bullying, but also different forms, their intensity, frequency and duration, location
where bullying takes place most often, perception of the phenomenon and response
towards the problem by all stakeholders (i.e. students, parents, school staff, educationists
12
and counselors). Several approaches and techniques are available to measure bullying
observation etc (Cornell, Sheras, & Cole, 2006; Crothers & Levinson, 2004). Following
Self-reports
bullying and victimization (Austin & Joseph, 1996; Borg, 1999) that provides precise and
either anonymous or confidential and this secrecy enhances the reliability of information
Moreover, most of the victims do not share their experiences with elders such as parents
or school staff, and bullying is not always overt, thus making it difficult to observe
(Menesi et al., 1997). Hence self-report measure serve as a useful source of accurate
violence to avoid social stigma while targets may either over estimate or deny their role
that children portray themselves pro-socially, for example as defenders of victims than
any other roles (Menesini & Gini, 2000; Sutton & Smith, 1999). Despite a few draw
13
backs, self-report methods is preferred by researchers over other methods due to its
Direct Observations
personally and/or by using audio/video recording techniques to explore the nature and
extent of bullying (Menesini, Melan, & Pignatti, 2000). Direct observations of bullying
permit an objective inspection of the child which is not affected by the internal group
of bullying and failing to come across the repetitive extent of behavior due to time
Teacher Reports
can provide valuable information about their involvement in bullying. Studies have found
bullying that are indirect and occur at place having lesser adult supervision than
classrooms (Hazler, Miller, Carney, &Green, 2001; Hunt, 2007; Pellegrini & Bartini,
2000; Shafer et al., 2002). Additionally reasonable convergence has been found between
students’ self-reports and teacher nominations of bullying and victimization (Beran &
Stewart, 2008).
14
Peer Reports
Peer reports are obtained through peer nomination method that affords larger
judgments and offer minimum risk of bias due to multiple raters (Salmivalli et al., 1996).
Children have greater opportunity to observe bullying directly and thus may provide
reliable information on the bullying roles of their peers. Still there are certain limitations
to this method for example; their evaluation can be subjected to be influenced by peer
reputation or their own relationship with the individual being judged. Furthermore, their
opinion may reflect inconsistency due to their moods or some specific event (McNeilly-
Choque, Hart, Robinson, Nelson, & Olsen, 1996). Risk of stigmatization can also
Guerra & Kim, (2010) found that peer reports generated lower bully and victim rates in
comparison with self or teacher/parent report yet no variation was observed for
be the only method to which all the others (i.e. observations and self-reports) are
Furlong, Sharkey, Felix, Tanigawa, and Green (2010) have highlighted the
insights into the occurrences of bullying and may deal with the disadvantages of a
Prevalence of Bullying
consistent finding across studies shows that a considerable number of students are bullied
by their peers at schools (Stassen & Berger, 2007). The initial studies were mainly
confined to Norway, (Olweus, 1973, 1978, 1991). It soon caught attention by researchers
in UK (Boulton & Underwood, 1992; Rivers & Smith, 1994; Whitney & Smith, 1993),
USA (Perry, Kusel, & Perry, 1988; Rodkin, Farmer, Pearl, & Van Acker, 2000; Nansel
et al., 2001), the Netherlands (Rutter, 1986; Junger, 1990), Australia (Rigby & Slee,
1991), Canada (Pepler, Craig, Zeigler, & Charach, 1993) and Japan (Morita, 1984; Taki,
2001a). Most of the findings are thus based on researches conducted in America and
Western European countries and a lot of variation had been found in occurrences of
bullying ranging from 18% (Carlyle & Steinman, 2007) to 82% (Gropper & Froschl,
1999). The official research initiated by the European Council in 2008 has acknowledged
that an average of 32% victimization rate varying between 25 to 45 percent from country
to country (as cited in Kukučionytė, Drąsutė, & Drąsutis, 2012). Cook et al. (2010)
reviewed 82 studies conducted in Europe, U.S and other parts of the world. The data was
23% as being bullied and only 8% of bully-victims. Depending on the nature and design
of the study, the participants’ age and the region, the proportion of school children
repeatedly victimized ranges from 7.1 %to 70.2 % (Due & Holstein, 2008).
For overall prevalence, we can say that over 20 % children worldwide continue
forms are concerned, studies have confirmed that direct physical and direct verbal are the
16
most frequently occurring types of bullying during elementary and middle school years,
while a significant number of students had been using or experiencing relational means of
bullying (Vettenburg, 1999; Martinez-Criado, 2014). Physical bullying had also been the
most frequent type at many places for example, Portugal (Pereira et al., 1996), and Japan
(Taki, 2001). Cyber bullying has emerged as a subtle form having high prevalence rates
among high school students (Ybarra & Mitchell, 2004). Recently, Modecki, Minchin,
Herbaugh, Guerra, and Runions (2014) found higher mean prevalence rates for traditional
bullying (35%) compared with cyber bullying (15%). The results were reported across 80
studies and the ratios were almost similar for both bullying and victimization. Further
Different types of bullying takes place at different grade levels. It has been
observed that physical bullying mostly occurs during elementary school years and verbal
and relational bullying occur in middle and high school years most often (Olweus, 1993;
Seals & Young, 2003). Researchers agree that victimization is peaked during primary
school, especially in grades four to six and that direct bullying rates decline as children
grow older and become mature to use more manipulative strategies such as indirect
bullying (Selekman & Vessey, 2004). This marked decline is usually associated with self-
reported victimization (Juvonen, Nishina, & Graham, 2000; Salmivalli, 2002) and hence
can be attributed to the egocentric nature of younger children who are more concerned
with what happens to them (Leff, Kupersmidt, Patterson, & Power, 1999). Most of the
17
targets are victimized by same aged peers studying in their own class and fewer had been
bullied by older pupils studying in higher grades (Zindi, 1994). Considering the high risk
of bullying involvement during elementary grades, the present study focused on children
that bullying others is more frequently reported by males as compared to females who
report more victimization (Seals & Young, 2003). Results of a study by Currie et al.
(2008) could not verify the gender difference in prevalence of victimization. In Asian
countries more males than females have been found to be involved in bullying behaviors
(Kim, Koh, & Leventhal, 2004; 2005; Kshirsagar, Agarwal, & Bavedekar, 2007).
Gender-wise prevalence of different forms of bullying has indicated that boys are more
physically abusive and use direct mode for bullying compared to girls. Although some of
the studies could not confirm gender differences in relational bullying (Crick & Bigbee,
1998; Prinstein, Boergers, & Vernberg, 2001) some others found that girls were more
engaged in indirect means while bullying their peers (National Crime Victimization
All kinds of bullying occur in a variety of contexts and usually last over an
extended period of time. Physical, verbal, relational, and sexual bullying might happen
18
over an extended period of time across different contexts, including the classroom,
hallways, playground, or traveling to and from school. Further, cyber bullying prolongs
ahead of the school day, through the use internet and digital devices.
(23.4%), during gym class (19.5%), and even in bathrooms (12.2%). Bowen and Holcom
(2010) reported that primary school students were most often bullied in school yards
(55%) and class rooms (24%), while other places where students were bullied included
corridors (4%), bathrooms (13%) and cafeteria (12%). Cyber bullying rarely happens
Duration of bullying has been associated with frequency of the behavior and the
outcome is more devastating (Sharp, Thompson & Arora, 2000). Short term bullying
usually includes indirect strategies, whereas long term bullying has been related to direct
forms especially verbal bullying. It can be concluded that nearly half of the victims are
bullied for some days or up to one months and a substantial number of them are bullied
for several months with a few experience bullying for years (Smith, 2014).
especially their parents, yet a considerable number of parents remain unaware of the
issue. Teachers are the second choice of students while reporting being bullied (Fekkes,
Pijpers, & Verloove-Vanhorick, 2005). Unnever and Cornell (2004) found that almost
40% students do not share their victimization experience with adults. That is why, fewer
consensuses have been found between teachers and parent reports of bullying and
19
students’ self-reports. Social support can play an important role in this regard inclining
victim to report their experiences (Holt, Kantor, & Finkelhor, 2009). The possibility of
reporting bullying experiences to parents reduces with age (Limber, Olweus, & Wang,
2012). It can be concluded that estimates of location and duration of bullying are
program in schools.
Assessment of attitude towards bullying was first emphasized by Rigby and Slee
(1991) who later developed Attitude to Victim Scale (Rigby, 1997). Since then, pro-
victim and pro-bully attitude has been studied in variety of context and a number of
bullying. Their understanding, insight and attitude affect the way they address the issue
(Poulou & Norwich, 2002). Teachers sometimes consider it difficult to deal with such
issues and try to avoid (Tomal, 1998), yet those who are able to relate the psychosocial
well-being of the students to their learning and academic performance are more likely to
intervene (Siris & Osterman, 2004) and acknowledge their responsibility according to the
Studies have been unable to find strong evidence on relationship between parental
attitude to bullying and children’s behavior (Eslea & Smith, 2000; Hanif, Nadeem &
Tariq, 2011). Parents’ attitude is reflected in ways they discuss the problem with their
children.It has been observed that parents either would not chat about such issues with
20
their children or often talk about being bullied than about bullying others. Moreover,
there is some evidence that parents of targeted female students were more likely than
parents of targeted boys to have a discussion with them about their experiences
(Houndoumadi & Pateraki, 2001). As far as parents’ active intervening role after being
2005) of elementary school students suggests that intervening the situation often results
in reducing the frequency of bullying, or worsen it. Hence, it can be suggested that
teachers’ understanding and seriousness about the problem and vigilant adult supervision
other psychological aspects of individual have been examined within broader contexts
and the interaction has been well explained in terms of social-ecology of bullying
(Swearer, Espelage, & Nepolitano, 2009). Cognitive explanation of bullying behavior and
its affective domain is highlighted The context of peer interactions pertaining to bullying
dilemma has been discussed in the light of theories such as dominance theory by
Pellegrini (2002), attraction theory of Bukowski, Sippola, and Newcomb (2000) and the
homophily hypothesis discussed by Cairns and Cairns (1994) and Espelage, Holt, and
Henkel (2003).
21
Dominance Theory
and control over weaker individuals. Pellegrini (2002) has argued about the need to
redefine the domination and assertiveness status by students within new peer circles
formed as a result of transition to middle school. This is one explanation of peak bullying
experiences during early adolescence. The social dominance viewpoint therefore suggests
that individuals are arranged into a hierarchy where everyone is placed according to the
varying access to available resources such as number of friends and individual standing
amongst them. Pellegrini and Burndt (2001) have speculated that using aggressive
strategies may ascertain dominance within peer relationship and the peer group then
Attraction Theory
rather than remaining dependent on parents. Attraction theory claims that these
youngsters urge to affiliate with age fellows who seem to enjoy more liberty and reflect
greater autonomy (e.g. showing aggression and assertiveness), and less complaisant
attitude (e.g., conforming, helping behavior; Bukowski et al., 2000; Moffitt, 1993).
Bukowski and colleagues (2000) used play characters in their research and identified that
children’s attraction to aggressive students increased upon the transition to middle school,
while appeal to peers who had high classroom competency (e. g. noble and pro-social)
from appeal towards an aggressive peer in order to gain similar status and power.
22
Homophily Hypothesis
individuals for example “homophily,” (Berndt, 1982; Kandel, 1978). These common
attributes may include demographic and behavioral characteristics (e.g., gender, race,
intellectual focus, misbehavior). Cairns and Cairns (1994) found support for homophily
hypothesis in relation to peer interactions and violent behavior during early adolescence.
Peers tend to associate themselves with students having comparable levels of aggression,
and they eventually persuade each other. In a study by Espelage et al., (2003) overall peer
school term. The findings were consistent across gender. Although aggression is
considered as a stable tendency, yet its escalated amount during middle school years can
Social-ecology of Bullying
Among different theories pervading the literature, social- ecology of bullying has
individual traits and their broader social set up across early childhood and adolescence
(Swearer et al., 2006). Bullying and victimization are reciprocally affected by personal
characteristics of the student, family background and socialization practices, school, peer
23
previous experience of victimization may place children at risk to bullying, but the
ratification of behavior may subject to the context and situation that consequently
promote or restrain such behavior (Espelage & De La Rue, 2011). The ecological
perspective not only considers the immediate developmental scenario, such as the
emotional environment provided by parents or teachers, but also broader context, that
may not include contributors the directly yet exerts an influence on them (Barboza et al.,
2009).
addressing cognitive aspects of bullying are linked to the findings from researches on
impulsivity and hyperactivity have been related to brain dysfunction (i.e. frontal lobe
injuries) (Liu & Wuerker, 2005). Segun (2004) has also speculated that the structures
associated with violent and aggressive behaviors could also relate to bullying behavior.
The aspect of cognitive functioning in bullying has been discussed in terms Social
Information Processing (SIP) model (Crick & Dodge, 1994) and the Theory of Mind
(ToM) skills (Sutton, Smith, & Swettenham, 1999; Sutton & Keogh, 2001). Both models
differ in their view point regarding the interpretation of social information by aggressive
youth. While Crick and Dodge (1994) claimed for the social skill deficit related to
24
aggression and bullying, Sutton and colleagues (1999) supported the notion of
processing bias and impairment in some of the stages proposed in the model (See Figure
1.1). The first step includes encoding of sensory information within a particular social
context into fragments of knowledge, which is interpreted by the child during second step
(2) followed by goal clarification and selection (step 3). In the next step (4), child seeks
the response options or creates it on account of its assumed value and appraisal of
available resource. After selecting the most suitable option (step 5), the child finally
capability to attribute mental states in terms of their own ideas, convictions, wishes and
intentions to themselves and others, and to employ that information to envisage and
comprehend behavior. Therefore, this approach proposes that many of the bullies at least
may have superior social skills rather than inadequate, provided that they are capable of
manipulating their peers and achieve their endeavor. Rather than considering all bullies
as incompetent in social skills and abilities, it has been proposed that several bullies
might be more accurate in perceiving their social worlds, and implement the facts for
their private benefits. This might be pertinent in the case of relational bullying, that
social information processing (as cited in McMahon, Felix, Halpert, & Petropoulous,
25
2009) has emphasized the role of normative beliefs and attitudes that serve as a filter
community violence can lead to development of more vengeful attitude and malicious
beliefs sustaining aggression in youth. Such circumstances result in developing low self-
efficacy and poor locus of control in turn leading to elevated levels of aggression
(McMahon et al., 2009). Findings suggest that not only cognitive discrepancies but
normative beliefs and attitudes also play an important role in understanding social
behaviors within environmental context. The view appears to support the social
understand and opt for responses in a manner that enhances the probability of
In addition, the SIP model also submits the role of emotions, even though these
were not included in the earlier model (Crick & Dodge, 1994; Dodge, 1986) but
behavior and attempted to merge them within socio-ecological frame work by examining
Liu and Graves (2011) evaluated major theoretical constructs addressing the
correlates of bullying and grouped certain theories into single perspective. The peer
context has been discussed within ethological perspective. The ethological structure
observes bullying from the standpoint of its advantageousness, such as in the course of
26
powerful and it is characterized by both negative (e.g., fighting, showing aggression and
bullying peers) and positive behaviors (e.g., leadership, guidance, compromise, center of
attention). Such behaviors are used by persons to struggle for esteemed social resources
theory. Affiliation also relates to dominance at initial stages of group formation. Thus the
perspective also incorporates the standpoint offered by attraction theory and homophile
premise.
The authors have argued about the role of autonomic patterns explaining
aggressive behavior among youth and highlighted the role of genetic and biological
environmental influences and hormonal imbalance are identified as risk factors related to
bullying. Liu and Graves (2011) considered ecological approach as not only an
processing models explaining aggressive behavior and its neurological correlates. The
mediating role of normative beliefs while interpreting social behavior in the light of
cognitive abilities has also been discussed by the authors within the context of bullying.
global framework of peer harassment that mainly takes into account the two well-
developmental model have been merged along with certain other theoretical explanations
27
of bullying and victimization. The reciprocal interaction and impact of different variables
Starting from individual dynamics of victims, active victims and bullies, the
authors move to the socialization agents and explain the aspects of familial and peer
group processes associated with bullying. This micro system incorporates specific
theories such as victimization has been explained in terms of learned helplessness, and
bullying with reference to the elements of social cognition such as shame management,
theory of mind, social information processing (SIP) and moral disengagement. Active
victims have been portrayed in the light of frustration-aggression process, theory of mind
and SIP. Notions of social learning and attachment theories better describe the family;
whereas peer group influences have been illustrated by social identity, dominance and
diffusion of responsibility.
Meso-system, the second part of the integrative model demonstrates the social
networks and school related aspects of bullying and victimization. Finally the macro
system has been explained with reference to the social groups and attitudes influenced by
social identity process and the broader effect of mass media described in terms of social
learning viewpoint.
The model best describes the relevance of each theory to specific stage and role
in bullying process. It also provides an insight to the interplay of various systems and
addresses all the stakeholders directly or indirectly related to peer harassment. Cultural
aspect should be taken into account while interpreting and employing the model to
empirical research.
28
Individual Characteristics
findings from twin studies have found low support for it and created controversies
(Brendgren et al., 2008, 2011). The mechanisms of influence may operate through
temperaments, cognitive and affective capabilities and patterns of social interaction (Ball
et al., 2008; Farrington & Baldry, 2010). At individual level, children’s physical and
psychological disability and academic weakness may also be important reasons for which
Other individual factors may include their physical appearance such as color or
the clothes they wear, temperament and certain characteristics of personality. Perry,
Hodges & Egan (2001) discovered an association between poor self-concept and risk of
being bullied. These children have low levels of self-esteem, competence and find
themselves helpless in critical situations like bullying incident. Individual traits interact
with other factors such as family, environment, and community; that can be better
Family
creates a sense of insecurity, dependence and lack of confidence and these characteristics
adhere to victimization experience. Contrarily, children who bully others may belong to
families who lack warmth and affection, have conflicting disciplinary practices and are
29
engaged in hostile and violent behavior. The bully-victim group may be nurtured in a
distressed or abusive family set up (Nickerson, Mele, & Osborne-Oliver, 2010). Over
criticizing the child, maltreatment and aggressive domestic atmosphere are also linked
with involvement in bullying (Holt et al., 2009).Parental negligence, lack of interest and
bullying as a perpetrator (Flouri & Buchanan, 2003). Lack of adult supervision during
childhood and adolescence can increase the risk of bullying (Espelage, Bosworth, &
Simon, 2000).
With respect to family structure and income, whereas school children in England
who belonged to non-intact families reported more bullying there was no association
between earnings and bullying (Flouri & Buchanan, 2003). Larger family size is also
considered an important correlate of bullying. Children living with more siblings are
bullying suggesting the similar distribution of bullies and victims across all level. On the
other hand, In Netherlands family’s poor socioeconomic status contributed more to the
School
School factors are characterized by school climate, school rules and policies and
Quenelle, 2010).These factors exert a significant impact on the student and the learning
30
environment. Benbenishty and Astor (2005) have presented a heuristic model addressing
the broader context of school violence and considered school related variables as central
school administration and rigid attitude of teachers also encourage violence at school.
Structural features and physical environment of the school such as school size, class room
conditions, teacher-pupil ratio and surrounding community are also associated with
Lack of interest by teacher in efforts to stop bullying has been widely documented
by researchers (Dupper & Meyer-Adams, 2002; Houndoumadi & Pateraki, 2001). This
belief is strengthened by reports on perception of students who mostly find the situation
hopeless and consider that teacher and their peers do not seriously attempt to stop
bullying at school (Adair, Dixon, Moore, & Sutherland, 2000; Unnever & Cornell, 2003).
school staff and administration turn a blind eye and overlook the incidents of school
bullying, the students perceive the aggressive behavior as tolerable and ordinary.
Lack of social support at school such as fewer numbers of close and trust worthy
friends or having friends who are not weaker themselves can be a significant risk factor
of victimization. It has been observed that social rejection and ostracisms contributes a
great deal in being bullied by peers. Salmvalli (2010) found that bullying is sometimes
31
aimed to enhance one’s popularity and social domination. Bystanders are important social
agents in bullying process. Those who have more pro-social behaviors like empathy and
also possess good social reputation can serve as effective and active defenders of victims
(Nickerson, Mele, & Princiotta, 2008; Caravita et al., 2009). Bullying may also result
from social identity process including in-group recognition and confirming to group
norms (Jones, Haslam, York, & Ryan, 2008; Jones, Livingstone, & Manstead, 2011,
2012).
media portrayal and economic inequality are some of the major reasons promoting
hostility, and other externalizing behaviors in children and adolescents have been related
to community demographics (Coulton, Korbin, Su, & Chow, 1995; Jonson-Reid, 1998;
Plybon & Kliewer, 2001). Researchers to study society and community related variables
in the context of school bullying are sparse (Cox, 1997). A Swedish study of middle
school pupils identified that perpetrators of school bullying were equally involved in
community level can also lead to involvement in bullying at school (Andershed, Kerr, &
Stattin, 2001).
mixed outcomes. While a Whitney and Smith (1993) confirmed a strong inverse
relationship between these two variables, Sourander, Helstelä, & Helenius, (2000) failed
32
to demonstrate any significant link between bullying and SES. A recent meta- analysis of
studies in US during past two decades concluded that adolescents belonging to poor
socioeconomic status were more likely to experience severe kinds of bullying (Fu, Land,
and social skills that increase the likelihood of being a bully or a victim. Keeping in view
the bullying associated with social and psychological difficulties (Alikasifoglu et al.,
2007), it is essential to study these aspects in relation to the theoretical approaches which
map the underlying mechanisms and elucidate the process of bullying. Therefore, we
shall have a closer look on the following important dimensions by that could contribute to
bullying.
1. Affective Processes
2. Peer Interactions
Affect is a broad term covering a wide range of feelings that individuals may
experience. It’s an inclusive concept that comprises both emotions and moods (George,
1996). Emotions are profound feelings that are aimed at someone or something (Farijda,
1993). Emotions are usually associated with particular situation and are considered more
transient than moods. Additionally, emotions are more vivid and expression is revealed
by non-verbal behavior e.g. facial expression Moods are feelings likely to be less
33
powerful than emotions and mostly occur without a contextual spur (Weiss &
Cropanzano, 1996). Moods are regarded to be more cognitive and may impel us to
ponder or speculate before taking an action, whereas emotion are thought to be more
action-oriented leading towards prompt response (Ekman & Davidson, 1994). Studies
admit that emotions and moods are mutually influenced by each other.
Reber and Reber (2001) defined emotions as specific labels referring to the affect-
laden states and the process of assigning them to a distinct category correctly from
recognition tasks mainly focus on response accuracy in identifying labels and associated
time to assess performance of the respondents. Researchers may use different sensory
modalities (e.g. vision, audition), yet most of the studies have examined non-verbal cues
cognition can be potentially mediated by affective processes (Lemerise & Arsenio, 2000).
Lemerise and Arsenio (2000) incorporated emotions into each step of the SIP model. The
emotions articulated by other peers, for example, may persuade the encoding (step 1) and
interpretation (step 2) of social indications, in addition to mood, which also affects the
other steps. People not only vary in encoding and expressing their own emotions, but also
understanding and decoding the emotions of other people and such variations have been
linked to social competence (Burks, Laird, Dodge, Pettit, & Bates, 1999). Moreover,
children’s mood states may also affect the interpretation of social situations and they
with certain mood may incorrectly identify the emotions of their peers thus increasing the
differentiate between a number of negative and positive emotions (Ekman & Oster,
1979). The cognitive stance of emotion recognition argues the existence of contextual
and social influences (Fox, 2004). In order to understand aggression in children aside
from social skill deficit, we shall have to look more systematically into the role of
Researchers have not only studied the disruptive impacts of emotions, but also
(Eisenberg & Fabes, 1992; Lemerise & Arsenio, 2000). Blair (1999) suggested that,
high levels of psychopathy in children and adolescents was associated with emotion
recognition deficits specifically sadness and fear. These two emotions are considered to
facilitate aggression inhibition. Other investigations demonstrated that those who are
abandoned ignored and named aggressive by their peers made more errors in recognizing
others’ emotions and interpreting relevant contextual precursors of the situation than
popular children (Bilello, Casiglia, Lo Coco, & Miceli, 1995; Ciucci, Tomada, De
especially with reference to relational bullying and reactive proactive aggression (Arsenio
35
and Lemerise, 2001; Camodeca, Goossens, Schuengel, & Terwogt, 2003). Studies have
(Woods, Wolke, Nowicki, & Hall, 2009), yet there is controversy regarding bully/victim
groups. Like aggressive and delinquent children, bullies might have an inability to
perceive the expressions of other student’s emotions accurately that could result in
violent behavior. Using the framework offered by Sutton et al. (1999) we can expect
hand might not be able to interpret the social/emotional cues in a threatening situation
and thus remain unable to avoid or defend themselves. Such tendency is likely to be
thrived by pre-existing mood state which may hamper the cue encoding and interpretation
stage of SIP (Schmid & Schmid Mast, 2010). Therefore studying mood states might have
groups.
It has been suggested that 'mood congruence' takes place when an individual
displays a positive association between his or her mood and the variable measured;
basically, any increase or decrease in mood strength reflects into the change in the
mood and the outcome variable; consequently, mood elevation results in decrease in
performance and vice versa. Mood cognition agenda (Bower, 1987) accentuates that
moods. Consequently, sad mood amplifies negative cognitions regarding the self, brings
about negative world view, and thereby turn out in persistent negative affect (Beck,
1967).
Active engagement in certain mood and affect laden information persuades the
evaluation process finally results in shaping the judgments and decision making (Forgas,
1995). The phenomenon is called affect infusion model (AIM) that has been studied in
the context of experimental mood induction and its significant impact on social cognition
in different age groups. It suggests that positive moods are linked to top-down processing
strategies (Gasper, 2003); negative moods, on contrary, are related to more in-depth,
positive words when induced with happy mood, yet there has been less support to mood
related attentional bias for negative induction. Although Bradley, Mogg, & Lee, (1997)
provided evidence for mood congruency effect under negative mood induction, other
studies could not demonstrate similar effect (Chepenik, Cornew, & Farah, 2007;
Gallardo, Baños, Belloch, & Rupierez, 1999; McCabe, Gotlib, & Martin, 2000). There is
also a tendency to label indistinct facial expressions as sad when individuals are in sad
mood (Bouhyys, Bloem, & Groothuis, 1995; Lee, Ng, Tang, & Chann, 2008). Positive
expressions; the recognition of mood congruent expressions was not affected by moods.
37
To study the effect of mood on any of the stages of Social information Processing
(SIP), an effective mood induction technique and reliable procedure is required. The
The trajectories of children’s affective development have shown that the skill to
Terwogt, 1984, 1986; Meerum Terwogt, & Olthof, 1989). Hence it can be supposed that
the ability to counteract negative mood outcome will develop consequently. Therefore, a
year old, while the strength of a positive mood bias is not affected by age (as cited in
children’s emotions and mood induction procedures (Erber & Erber, 1994; Harper,
Lemerise, & Caverly, 2010). Several mood induction procedures exist for inducing mood
should be more cautious and careful while manipulating mood among child samples.
Methods of mood induction stab to generate an emotion that is closer to the original
38
within the laboratory set up. Researchers have speculated that certain provoked moods
such as happy, sad, and neutral are suitable analogues of naturally-arousing emotions in
children (Barden, Garber, Duncan, &Masters, 1981). Even though a few researchers
suggested that emotions in induced form are not as much of deep and strong than
emotions that occur naturally, induced emotions are still considerably significant (Martin,
1990). Despite mixed findings regarding the efficacy of mood induction procedures
among children, it is yet very important to explore emotions and mood states in young
ones during their middle childhood and early adolescence because of the importance of
children using experimenter behavior and instructions, guided imagery, written and
verbal scripts, suggestive videos, feedback about performance, material rewards, musical
excerpts, autobiographical recall, social situations and watching faces. Video-based mood
convenient use and enduring outcomes. These methods were found to be equally useful
for inducing certain positive and negative affect states such as happiness, sadness, anger
and fear. Moreover, these techniques minimize the risk of demand characteristics due to
Authors have suggested that effects of pleasant mood manipulations is more long
lasting than unpleasant ones and therefore should be conducted at the end (Carlson &
Masters, 1986). Furthermore, experimenter should address and effectively deal with the
another important issue while dealing with child samples. A variety of procedures are
39
available, some of which assess the dimensional aspect of affect (e.g. positive and
negative), and others measure the specific type of emotion intended to induce (e.g.
happiness, sadness). It has been observed that simple forced choice task yield significant
and consistent results for children as compared to adults (Kratus, 1993). Dolgin and
Adelson (1990) found considerable variations when more alternatives were offered, and
when the responses were evaluated on the basis of accuracy. Both visual analogue scales
and verbal response format have been used by researchers as mood manipulation checks
Peer Interactions
(Bagwell, 2004; Dunn, 2004; Schneider, 2000). Children’s activities and practices during
peer interactions have been associated with their social skills and competence (Zeller,
Vannatta, Schaffer, & Noll, 2003) that mostly result in greater adjustment in family
relationships, attachment and intimacy (Doğan, Karaman, & Çoban, 2012; Field,
Diegoi, & Sanders, 2002), academic performance (Rabaglietti & Ciairano, 2008),
personal well-being and satisfaction and success (Çevik Büyükşahin, 2007; Keefe &
construct at dyadic level (Bukowski, Newcomb, & Hartup, 1996). Friendship is a close,
reciprocal and voluntary bond and affection that is mainly characterized by positive
features (Berndt, 2002; Schmidt & Bagwell, 2007). In his influential effort on studying
relationships throughout the lifespan, Sullivan (1953) used the term chumpkins to
during middle childhood and early adolescence (Berndt, 2004). Children become more
aware and conscious about long-term relationships during this phase of life (Parker &
Seal, 1996).
quality can be determined by certain inter linked positive (e.g., care, companionship,
conflict resolution) characteristics and a few negative features for example conflict,
jealousy, deception, and rivalry (Parker & Asher, 1993; Berndt, 1998; Rubin, Bukowski
& Laursen, 2009) associated with it. Yet, a high quality friendship incorporates more
positive features; friendships are thought to be high in quality when they have high levels
of positive features and low levels of negative features (Berndt, 2002; Schmidt &
Bagwell, 2007). Berndt (2002) further discussed that actually positive features contribute
to high and low dimension of friendship quality. On the other hand, overlapping in
positive features may affect the assessment of friendship quality as argued by Thien and
Nordin (2012).
41
Friendship often serves the purpose of being a pro protective buffer against
various negative factors that occur in individual’s life. Thereby, it is considered as one of
the factors that can prevent the process of bullying. Individuals that have low level
quality of friendship express tendencies towards both bullying and victimization (Berndt,
2002). In this case, the quality of friendship is a determining factor towards its protective
capacity. If the children involved in the process of bullying do not have friends that
would support them, they are more likely to be victimized (Bollmer, Milich, Harris, &
Maras, 2005). On the other hand, if there are friends that are able to support the victim or
There are also many studies that classify the victimized children as the group that
other groups involved in bullying behavior (Didaskalou, Andreou, & Vlachou, 2009;
Forero, 2013; Juvonen et al., 2003; Storch, Brassard, & Masia-Warner, 2003). According
to other findings, bullies are more socially competent in making friends, and report lesser
seclusion or rejection (Nansel et al., 2001). Whether the friendship quality is influenced
The researches show that the children with high quality friendship are less likely
of the patterns of behavior, the existence of friendship positively influences one’s social
interactions (Berndt, 2002). However, if the person expresses the qualities associated
with bullying behavior, they are less likely to obtain the high-quality friendship.
42
processes and patterns of peer interactions influence the short and long term emotional
overall wellbeing.
Consequences/Effects of Bullying
personal level. These consequences of bullying are deleterious and touch not only
offenders and victims (Gini & Pozzoli, 2009) but also the bystanders (Nishina &
Juvonen, 2005). Nevertheless, the toll is greater for those who are directly involved in it
as offenders and sufferers than those who are indirectly associated with bullying or
remain uninvolved. At larger level, negative outcomes of bullying influence the school‘s
quality, learning and social environment (Osterman, Björkqvist, & Kaukiaine, 1998) and
contribute to the prevalence of criminal behavior in society. Long- term effects include
mental health issues in adulthood (Smokowiski & Kopasz, 2005; Ttofi, Farrington, Losel,
& Loeber, 2011), antisocial, legal convictions (Farrington, 1993; Olweus 1992).
Effects on Victims
poor academic performance, fewer number of close friends, have a cynical view of
school, face health issue both physical and psychological. They are also more prone
towards psychosomatic complaints such as sleep disturbances and bed wetting (Due, et
al., 2005). They usually remain upset and anxious about what happens to them at school
43
and fretful, alone, sad, depressed, insecure, reserve and overly compliant (Frost, 1991;
bullying incidence results in school avoidance and absenteeism that ultimately affects
their academic performance (Kochendrfer & Ladd, 1996; Juvonen et al, 2001). Low
achievement at school and poor performance are variedly associated with bullying. Some
researchers have linked it to victimization (Myerd & Joseph, 1997) while others have
found it within bully-victim group or the perpetrators of bullying (Nansel et al., 2001).
Victims also show poorer problem-solving skills than those who have not been bullied.
Effects on Bullies
smoking, drug use (Berthold & Hover, 2000; Nansel, et al., 2001) and other problems
mostly externalizing in nature. It has been found that involvement in bullying as offender
strengthened antisocial and violent attitude as school bullies were frequently involved in
criminal activities as adults (Olweus, 1992). Bullies are penalized for disciplinary issue,
may harm others outside the school as well engage in self-destructive behaviors (as cited
Bullies also face problems in school adjustment that affects their academic
disrupted academic and social life at school that may result in delinquency and conduct
44
problems in later life (Farrington &Welsh, 2005). As compared to victim and bully-
victim groups, bullies are less likely to experience internalizing problems, such as
Schonfeld and Gould (2007) demonstrated that perpetrators are as likely as maltreated
Studies have suggested that bully-victim group may show the poorest functioning
as they experience both internalizing and externalizing problems and health related issues
as compared to pure bullies or pure victims (Nansel, Craig, Overpeck, Saluja, & Ruan
2004; Veenstra et al., 2005). Nansel, Overpeck et al. (2001) have also demonstrated
similar kind of findings that students who experience and perpetrate bullying are more
with bullies. Wolke, William, Adrian, and Costello (2013) have discussed the devastating
helplessness and social deficits continue to influence the individual in adulthood as well.
Bully-victims may also suffer financially as adults as they are less consistent at their jobs
like pure bullies (Wolfe et al., 2001). Involvement in bullying at school as both offender
and target is associated with the probability of worst mental health outcomes and illegal
Researchers have asserted that bullying may result in academic failure, health
issues, loss of social relations and a variety of cognitive, affective and behavioral
45
problems (Duncan, 1999; Sullivan, 2000; Selekman & Vessey, 2004). A number of
factors have been associated with bullying and its consequences. In fact certain
individual, family, school, and societal aspects have been found to interact reciprocally in
since the academic toll of both being bullied and bullying others is high (Juvonen, Wang,
defined academic competence has been defined as skills and capabilities considered
achievement (i.e. grade point average). High levels of perceived social support from
family and peer has been linked to higher grades in exams (Ma, Phelps, Lerner, & Lerner,
2009; Rosenfeld, Richman, & Bowen, 2000).Research has also demonstrated that the
majority of students, who have positive peer interactions and intimate dyadic
2004). This salience of peer factors in academic performance of students is evident and
requires further exploring of the context and elements of this relationship. Breaking it
down to positive and negative dimensions, both quality of friendship and bullying has
been found to affect students’ academic performance (Erath, Flanagan, & Bierman, 2008;
Juvonen, Nishina, & Graham, 2000). Veronneau and Dishion (2011) suggested that
students. Friendship provides a buffer against social and academic pressures (e.g.
victimization and poor performance). There is need to uncover the connections and the
present study aimed to fulfill that need studying both friendship quality and bullying and
Literature Review
typical part of childhood has now emerged as a serious social and ethical issue at schools
(Smith & Monks, 2008). According to APA database, researches dealing with the issue
of peer harassment and victimization could also be classified as bullying (as cited in
Postigo et al., 2013). Various terms have been used to describe the construct of
before studying other components of the phenomenon (Carrera et al., 2011; Lee,
2011).
Smith and Monks, (2008) have emphasized the cautious use of words referring to
bullying while translating the term in other languages that encompass the actual sense of
bullying. Inconsistencies have been found even across English speaking countries, for
UK. A term rarely equals bullying in its meaning when translated into another language
The word that represents bullying in a certain language corresponds to the nature
of bullying behaviors that actually occur in a particular culture. Linguistic differences can
be avoided by emphasizing observed behaviors in a translation process; yet may not prove
to be very useful in translating bullying. “Ijime” has received much attention most recently
as the Japanese term for bullying; although it emphasizes more on psychological and
While reviewing the nature and extent of bullying and victimization in South
Korean schools Koo, Kwak, and Smith (2008) have identified some distinctive features.
Unlike Western countries, most of the Korean school students were involved in bullying as
perpetrators and the role was mostly reported by girls as compared to boys. The construct
of bullying named as ‘Wang ta’ in Korea was found to be less physical in nature, rather
verbal and relational forms were more common among students. For the sake of
bullying in some culture. Hence authors have emphasized to take the indigenous
perspective into account and conceptualize the phenomenon accordingly. The study used
senior version of Olweus Bully Victim Questionnaire after translation and adaptation for
Korean population of school aged children. With specific reference period and term Wang
Ta, Korean students were less likely to report victimization (5%) compared to previous
48
studies (24% and 30 %) (Parken et al., 1999). The concept of Wang Ta and the associated
findings were consistent with findings on bullying explained by the term Ijmi in closely
related culture of Japan (Morita et al, 1999). In the light of their findings, authors have
postulated that nature of bullying is more collective in Asian countries as a result; verbal
assaults and social exclusions are more prevalent in Korea and Japan.
Among several other researchers attempting to define bullying; Smith and Monks
(2008) reflected upon the concept of bullying in its developmental and cultural context.
By extensively reviewing the fact from its origin through various transitions and
purposeful, harmful, repetitive and unfair, including power differential, sense of pleasure
Prevalence estimates drawn from large scale national surveys in U.S during last 5
to ten years demonstrated various prevalence rates for school bullying, ranging from 11%
(Iannotti, 2012) to 20% (Eaton et al., 2012). Robers, Kemp, and Truman (2013) however
have established a rate of 28% as a result of School Crime Supplement reports. In another
study by Swearer, Espelage, Vailancourt, and Hymel (2010) prevalence estimates range
According to the most recent reports of U.S. Department of Health and Human
students in grades 6-12 purportedly have experienced bullying. More than 70% students
49
have reported witnessing bullying at their schools and nearly 30% youth confessed to be
Earlier studies in Canada, nearly 15% of children reported bullying others; while
fewer (5%) students were found to be victimized (5%) of children reported bullying
others in school at younger age level (Charach, Pepler, & Ziegler, 1995; Craig, Peters &
national sample found 5-13 % students as bullies and 9-21% as victims (Currie et al.,
2008).
While reporting the results from 40 European countries based on the Health
Behavior in School-Aged Children (HBSC) survey for year 2005-2006, Craig et al.
(2009) established prevalence estimates with a range of 8.6 % to 45.2 % for boys, and 4.8
% to 35.8 % for girls. In general, prevalence rates of bullying were lower in Scandinavian
intervention efforts especially the Olweus program. They observed a clear decline in
bullying among boys while age-related patterns were inconsistent for girls.
claimed that they had been bullied frequently at school during one month period
(Chamberlain, George, Golden, Walker & Benton, 2010). Jansen et al (2012) revealed a
more optimistic picture in the Netherlands by finding only 4% elementary school students
reported being bullied. On the other hand 17 % admitted themselves as bullies and 13%
reported as being bully victim. However, another recent finding from a large scale sample
representing all provinces, confirmed high victimization rates (31.8%) among students
50
(Oldenburg et al, 2014). Alikasifoglu, et al. (2004) found that almost 30% students
27%, 32%, and 20% respectively especially in elementary schools (Cross et al, 2009;
Rigby, 2007).
Research in African countries also present the a bit different picture with almost
45% reports of students who had been bullied at least once in a month. Restricting the
Kann, 2008). High rates of bullying (16-61%) have been reported in South African
middle and high schools by different investigators (Mlisa, Ward, Flisher, & Lombard,
2008; Neser, Ovens, van der Merwe, Morodi, & Ladikos, 2003; Townsend, Flisher,
Bullying prevalence estimates from 36 countries in East Asia and Pacific region
have been discussed in the section of peer violence in a systematic review of research on
ranging from 10.2% to 32.9 percent within probability samples whereas lower rates had
been reported within convenient samples (2.7% to 16.9%). Of these studies, 21 are
drawn from some type of probability sample and 12 are of school-based and convenience
samples with the remainder being qualitative studies. Victimization reports ranged from
5.8% to 35.5 % for probability samples and higher incidence (up to 70%) was recorded
within convenient samples. In addition the highest estimates of peer violence (70 % to
Western countries. This region mostly comprises collectivistic culture and wide range of
diversity has been found for example admittance to the role of a perpetrator varied from
10.2% in Korea to nearly 33% in Thailand (Laeheem & Baka, 2009). Children who
reported themselves as targets of bullying were only 5.8% in Korea (Koo et al., 2008) and
35.5% in Philippines (Rudatsikira & Siziya, 2008). Kim, et al. (2004) had demonstrated a
bit higher rates for bullying (17%) and victimization (14%) in Korean schools. Around
20% Chinese middle school students experienced peer victimization during a period of 30
days (Hazemba, Siziya, Muula, & Rudatsikira, 2008). Japanese students are at greater
risk of bullying and victimization compared with American or South African students
found lesser peer victimization rates among Indian students (25-31%) compared to the
(2007) established the same trend in Indian schools (31%) with more reports of verbal
bullying followed by physical from. Results from a survey by World Health Organization
have depicted the worst picture in Pakistani schools by reporting an overall prevalence of
42% (Sheikh, 2013). Another recent source has established a rate of almost 20 to 24% for
Statistics on bullying from different parts of the world indicate that bullying has
become a serious problem for school aged children that can be addressed by
Frequency and nature of bullying also varies according to children’s age and
grade level. Older children choose more subtle and indirect tactics for bullying others,
whereas younger children studying in elementary grades are more often involved in direct
forms of bullying (Craig et al., 2009). Bullying has usually been found to be most
common among middle school students (Nansel et al., 2001); however, studies have
recommended that bullying peaks during the time of school shifting such as elementary
to middle and middle to high school. Opposed to the findings of bullying peaked during
8th grade (Espelage, Holt, & Henkel, 2003), Olweus (1991) had reported higher incidence
among elementary grades. Reynolds (2003) found the bullying curve peaked during 6th,
8th and 12th grades and hypothesized these spikes to be associated with school transition
periods. This has been considered a critical period characterized by formation of new
groups bullying serves as a mean of achieving peer domination and control (Pellegrini et
al., 2011).
Being described as gendered concept (Rodkin & Berger, 2008), bullying needs to
be examined by underlying processes that account for the differences and similarities in a
given context. Studies found support for boys’ over-involvement in bullying perpetration
and victimization as compared to girls (Nansel et al., 2001; Pepler, Jiang, Craig &
Connolly, 2008). While arguing about such gender disparities, Underwood and Rosen
(2011) have emphasized to consider different forms of aggression used in the act of
bullying. Evidence clearly supported boys were overly involved in more direct forms of
aggression (verbal and physical) whereas the differences were less clear for relational
53
aggression (Card, Stucky, Sawalani, & Little, 2008). Hence the authors have elucidated
Underwood, 2003) comprising both direct and indirect, verbal and non-verbal dimensions
of relationship damage intended in such kind of bullying. They have discussed the role of
Underwood and Rosen (2011) have also analyzed how boys and girls conceive the
countries from Asia and Europe (Smith, Cowic, Olaffson, & Lefooghe, 2002). Children
are able to distinguish among various forms of bullying during middle school years. The
phenomenon. Conversely, Vaillancourt et al. (2008) found that social aggression was
highlighted within definitions of bullying provided by girls. The authors have attempted
to unfold the social processes underlying bullying while reporting the contradictions
found in gender literature on bullying. There had been evidence for girls remaining either
uninvolved or assuming the role of defender against bullying while boys mostly acted as
Hawkins, Pepler and Craig (2001) suggested that both boys and girls effectively
intervened within same gender discord. Yet the nature of intervention by girls was more
extended for cyber bullying reports that is considered a more subtle form of bullying.
54
Since the rates of cyber bullying using various technologies have risen during recent
years, it is important to discuss if the gender based relational bullying hypothesis sought
approval or rejection in these investigations. Having stronger verbal abilities, girls have
been reported to be involved in cyber bullying more often than boys (Hertz & David-
Ferdon, 2008; Hinduja & Patchins, 2009; Kowalski, Limber, & Agatston, 2008).
reports (Li, 2006; Raskauskas & Stoltz, 2007). It is truly the time to move beyond mere
averages as suggested by Underwood and Rosen (2011) for extending the investigations
to the core factors and course of action to explicate such inconsistencies that are more
likely to be attributed to conceptual and methodological issues. Finally, the authors have
status (SES) and risk of being bullied and suggested that these victims used to experience
more adverse psychosocial outcome as compared to the victims who belong to more
affluent social set up (Due, Damsgaard, Lund, & Holstein, 2009). Not only family’s
social status but the school neighborhood is also associated with bullying.
Jensen et al. (2012) explored how several indicators of SES are related to school
and type and earnings indicated family SES, while school neighborhood SES was
different types of bullying and victimization. Consistent with prior findings, results
55
showed association between family’s social disadvantages (e.g. parents’ younger age,
single status and lower level of education) and involvement in bullying perpetration.
impact of socioeconomic status in bullying research, present study has also investigated
Most of the studies rely on self-report measure of bullying that can be classified
into two categories: definition based questionnaires and behavior based questionnaires.
Behavior based self-reports do not use the word bullying in the questionnaires (Espelage
& Holt, 2001; Reynolds, 2003). On the other hand Olweus (1996) and Ma (2002) have
referenced the word bully or bullying in their measures. Researchers have argued the
potential issues related to the response pattern affected by social desirability among older
age groups while referring to the term bullying in self-report (Espelage & Holt, 2001).
Olweus and Solberg (2003) have mentioned that anonymity can prevent the students to
portray them in a socially desirable manner. Other researchers have also included the
term bullying and its definition while assessing the construct (Salmivalli et al., 1996).
The definition adds to the clarity and unanimous comprehension of the term bullying for
comprehensive self-report instrument for middle and high school students with strong
56
psychometric properties (Austin & Joseph, 1996; Kyriakides, Kaloyirou, & Lindsay,
2006; Olweus, 1997; Pellegrini, Bartini, & Brooks, 1999). It is equally applicable to
primary school students with clear instruction and cautions suggested by authors. It not
only covers the incidence of overall bullying and its forms but also investigates the
bullying, and general satisfaction with school (Olweus & Solberg, 2003). It introduces
participants. The word bullying, bullied and being bullied has been used several times in
Olweus Bully/Victim Questionnaire or a subset of questions from it and the measure has
been successfully standardized across different cultures (Schwartz et al., 2001). Whatever
measure we use, vocabulary and sentence structure should merit special considerations
Teacher nomination procedures are also helpful and Olweus (1993) regarded
these procedures as accurate source of information. The teacher is either asked to focus
on the students in the class and behavior ratings are obtained for pupils (Leff et al., 1999;
Monks, Smith, & Swettenham, 2003), or they have to assign the students in a list into
different bullying roles. Though pointing to the issue of personal bias in teacher ratings of
bullying behavior, Pellegrini and Bartini (2000) have confirmed the relationship between
teacher ratings and peer nominations. Adult (e.g. teacher) reports were found to be
associated with self- reports and peer nominations (Smith, 2004). Despite several
limitations, many researchers continue to use teacher reports for assessing bullying roles
57
particularly in junior school settings (Jenson et al., 2012; Perren & Alsaker, 2006). Multi-
method approach is suggested by experts to obtain more reliable findings about bully-
the cognitive and affective incapability precede the bullying? Regardless of this
depends on their ability to discern and understand other’s emotions, and the awareness of
their own emotional state. These abilities are also helpful in regulating interpersonal
exchanges (Saarni, 1990). Denham, von Salisch, Olthof, Kochanoff and Caverly (2002)
restricted to recognizing other’s emotions from verbal and non-verbal cues, but also
implies the awareness of antecedents and consequences of elicited emotion. This concept
is also an integral part of theory of mind and self-control (Dunn & Cutting, 1999). Both
emotion recognition and theory of mind skills are thought to be the main components of
social cognition. The present study focuses on the ability to recognize facial expressions
of emotions that provide the first signal of change in emotions and serves as a foundation
58
upon which is based the understanding of emotion and theory of mind skills (Denham,
1998).
expression of six basic emotions i.e. anger, fear, sadness, disgust and happiness (Ekman,
1973). There has been evidence that face expression recognition skills are subject to some
recognize happy faces is evident at the age of 7 months (Ludemann & Nelson, 1988).
Nelson (1987) found that infant can discriminate valence in emotions by 12 months of
age. Darwinian (1965) approach proposed that affection recognition ability from facial
cues emerges as infant begins to respond to mother’s smile or frown. These non verbal
gestures are the first indicators of social interaction at 3-5 weeks of age as the first social
smile appears (Izard, 2001). Reaching the ages of 3 to 5 years; children are capable of
recognizing original and posed facial expressions of emotions (Camras & Allison, 1985).
This understanding and awareness enhances upon child’s admission to school and he/she
becomes able to make assumptions about other’s judgment in the light of their prior
recognition arises prior to the onset of adolescence especially at 11 years of age (Tonks,
By this time, patterns of peer interactions also change ardently and children feel
more connected to the peer group than family. Children also form and understand the
59
need of intimate peer relationship and hence the friendship becomes a source of sharing
emotions with one another (Denham, Mitchell-Copeland, Strandberg, Auerbach, & Blair
1994). This stage of childhood has also been considered crucial for experiencing peer
victimization and displaying aggressive tendencies. Dunn and Cutting (1999) discovered
that emotion understanding in early childhood was positively associated with healthy and
less conflicting social interactions. Contrarily, children with low emotional competence
may experience disrupted social relationships and more likely to show negative and
hence the emotional component can be studied to explicate the underlying cognitive and
across deviant child behaviors (e.g. conduct disorders, anti social behavior) and
aggression being associated with emotion recognition deficit or advantage (Leist &
Dadds, 2009); however research pertaining to bullies and victims is scarce. Poor emotion
recognition ability may account for the poor social skills of students involved in bullying
recognition tasks, signifying the role of negative social relationships in biased affective
can help perpetrators to identify the targets’ weaknesses and thus they might use this
(Lemerise & Arsenio, 2000). This approach counters the typical model of social skill
deficit (Crick & Dodge, 1994), while confirms the explanation by Sutton (2001).
60
victims has been conducted by Woods, Wolke, Nowicki, and Hall (2009). Children’
ability to understand the affective component has been studied in relation to their
general, Woods et al. (2009) found highest error rates for recognizing angry faces and
least errors were found for recognition of happy faces. Consistent with the prior findings,
the study showed significant gender differences as girls’ outperformed boys. Results also
depicted that students who were subjected to relational bullying only and those who
experienced both relational and physical forms (overlap) consistently showed higher
mean errors in recognizing angry and fearful faces as compared to the neutral group and
bullies. Their general ability of emotion recognition was also poorer than neutrals. Study
could not demonstrate any relationship between empathy and bullying roles however
boys exhibited lower levels of empathy than girls. The results provide partial support to
misinterpretation of affective cues and incorrect judgments lead them to faulty behavior
Since fewer researches have directly examined the relation between bullying and
emotion recognition, it is likely to believe that poorer recognition ability may lead to
being victimized and likewise, that complications may occur as a result of being
persistently victimized. Wood et al. (2009) elucidated that the global recognition deficit,
especially for angry and fearful expression, of children experiencing relational bullying
might be one rationale for why they were frequently targeted by offenders. Mahady-
Wilton, Craig, and Pepler (2000) have also suggested the same that victims are at risk of
61
have reasoned that experience of social isolation as a result of rejection and victimization
prevents them to be involved in social interactions and thus leads to enhanced social
enhancement proposition argues that victimized children may possess better emotion
certain emotions is improved a result of negative peer interactions. Yet the later
17years old victimized and non-victimized. Overall accuracy was same for groups
whereas, recognition of sadness and anger was different across victimization status.
Victims were less accurate than non-victims. Thus the deficit hypothesis is usually
supported partially (i.e. for specific emotions) in researches. Irwin also added the aspect
of chronicity to the existing research framework with a sample of 10-14 years old early
responses across six emotion categories between non-victims and frequently victimized
students. However as the study used the previous paradigm of resenting emotions in
chronic victims in 2nd segment thus lending some support to enhancement hypothesis.
information processing (SIP) model. They suggested that affective processes including
62
moods and emotions can influence all the stages of SIP in distinctive manner i.e. from
cue encoding and interpretation to utilizing previous social knowledge in goal persuasion
and response decisions. Children get into the social scenario of school with a preceding
mood that might be irrelevant of that situation, such as anxiety about some domestic
affair. They also differ in ways they regulate their pre-existing emotions. Inability to
regulate such affective predispositions has been linked to SIP patterns associated with
higher levels of aggression among school children. Garaham, Hudley, and Williams
(1992) had also explained how mood states had influenced encoding and interpretation of
In a study by Orobio de Castrro et al. (2003) children having negative mood states
children were exposed to mild or moderate emotional arousal, they made more social
goals over relational goals more than their non-aggressive peers. The rejected-aggressive
children have been found to be more reactive to emotional signals and this reactivity may
influence their cognitive style. These results, combined with Eisenberg and Fabes (1992)
view about individual differences in children’s emotionality and the ability to interpret
other’s emotions, suggest that emotions are important component to integrate into SIP to
better understand social adjustment. This includes the emotions experienced by the actors
Schmid and Mast (2010) established the effect of mood on emotion recognition in
a sample of young adults (51 women and 42 men) having an average age of 23 years.
63
Film scenes were used for sad and happy mood priming, whereas a screen saver provided
the neutral condition. Subjects then performed an emotion recognition task using happy
and sad facial expressions of varying intensities. It is important to note that the authors
used to play happy and sad instrumental music during emotion recognition task
performance to sustain the effect of mood. Manipulation check showed efficacy of mood
priming in all the three conditions. Results of mixed model ANOVA supported the
typical gender bias in emotion recognition as women made lesser errors than men.
congruence found for participants in sad mood who recognized sad faces better than
happy faces. Mood incongruence was reported for happy condition. When others’
emotions are not consistent with our own mood states, we may find it difficult to identify
and understand their emotions and situation can lead to disturbed social interaction.
Similar findings have been reported by Lee et al. (2008), who studied the impact
of temporary mood states on emotion recognition among Chinese young adults. They
used cheerful film clip, a documentary on tsunami, and anti-smoking campaign for
inducing happy, sad and neutral mood respectively. The subjects went through emotion
happy or neutral mood conditions. Confirming the negative attention bias in emotion
recognition, authors have discussed the visual organization and attentional system
disturbances.
64
Contextual cues can lower or enhance the information processing rate and also
contribute to emotional arousal (Bower, 1981). Paying less attention to mood incongruent
information and misinterpreting others’ emotions have certain implications for peer
relationships in school setting. Emotion recognition skill has been found to be a precursor
to emotion regulation and the quality of social relationships. Existing findings in this
recognition abilities than non-victims. Studies have established link between negative
mood and elaborate processing (bottom up) during cognitive task performances (Deddin,
People in sad mood are more sensitive to small details as compared to those in
happy mood states who are shown to use top down approach and process the upshot of
the information (Gasper & Clare, 2002). Hills, Werno, and Lewis (2011) have discussed
about the reduced holistic processing and attention overload during sad moods. On the
other hand cognitive overload in happy mood states may hamper the recognition
recognition of affect. Nature of cognitive involvement and task complexity can affect the
aspect. Mood congruent attention bias was found for sad moods in simple tasks for
example old/new face recognition. Although participants in sad mood were more accurate
in recognizing all types of facial expressions, yet mood congruency was found for
recognition, they found minimal support for the effect of mood on performance in all the
65
three conditions. It suggests that certain confounding may influence the result such as
intentional learning strategies (Jerman, Van der Linden, & Agernbeaus, 2008).
Chepenik et al. (2007) have also found sad mood incongruence in affect
recognition accuracy. Schmid, Mast, Bombari, Mast, and Lobmaier (2011) have
emotions with an exception of happy face; thus supporting the typical happy face
(2004). Studies have established the relationship between attention bias to threat and
patterns of social and emotional functioning, mostly in the case of anxiety (Bar-Haim,
Haim, Nelson, & Fox, 2009). Pérez-Edgar et al. (2011) addressed this issue by representing
connections among temperament, attention bias to threat, and social functioning in a large
sample of 5 years old children. Measured during toddlerhood, behavior inhibition proved
moderated by attention bias to threat. There has also been evidence for attention bias in
include a prominent contribution by Rennels and Cummings (2013). They used affect
laden stories to induce positive and negative moods and a list of table manners produced
neutral condition. Results showed that children more correctly identified the emotions in
vignettes. Yet these differences were evident for younger children (5-7 years) compared
those entering their middle childhood (8 years old). Analysis demonstrated that provided
with the contextual enquiry, children’s accuracy was higher for negative emotions.
Furthermore, happy mood enhanced the response speed than negative mood, albeit the
accuracy was not affected by mood state. Overall findings elucidate that task complexity
Kelley, Liu, Rodger, Miellet, and Caldara (2011) recognized that people from Asian
countries (e.g. China) tend to use more holistic approach than people from Western
countries who are more likely to use deliberate processing. A cross-cultural sample of
children (7-12 years old) from England and China was concurrently assessed to explicate
the effectiveness of this claim. Findings revealed that children’s patterns of face
processing were similar to the adults in respective cultures. It was further suggested that
children mainly focus the internal features during face processing than external ones.
Given the support to attention bias and mood congruence in cognitive processes
investigate this endeavor with reference to school bullying and victimization. Observing
how children infer the expressive meaning of others' facial expressions gives information
for how children interpret the objectives and intentions of others in interpersonal
exchanges. The skill is essential for developing and sustaining healthy and smooth social
major concern and priority providing the individual a platform to address their identity
issues (Noller, Feeney, & Peterson, 2001). Peer interactions occupy 30% of child’s
socialization in middle childhood (Rubin, Bukowski, & Parker, 1998). Hartup and
Abecassis (2002) suggested that almost 85% children admit to have friends during middle
childhood. Making and sustaining the intimate peer relations entail a range of
proficiency, self-regulation, perceiving the intention, needs, and idea in others, social
information processing skills and problem solving approaches, among others. These
abilities develop and nurture during childhood, and nature and prospect of friendship
patterns are subject to developmental changes and become more stable and reciprocated
and contentiousness (Bukowski et al., 1996; Furman, 1996). Despite contentions about
potential bias in self-report, investigators agree that children’s personal insight and
opinion of their friendships may provide the best source of evaluation; because no one
else can better describe the extent of shared emotional acquaintance (Furman, 1996;
Smith & Brownell, 2003). Children are usually asked to rate how regularly specific kinds
68
of exchanges occur with a particular friend (e.g., conflict, pro-social behavior, and self-
(Parker & Asher, 1993). While Berndt (2002) has used two dimensional approaches
pertaining to positive and negative characteristics; Bukowski et al. (1994) have preferred
self-report. Friendship Quality Scale (FQS; Bukowski et al., 1994), the measure used to
Studies have indicated that closer friendship relation for example best friend may
exert greater impact on children’s adjustment as compared to other friends or the broad
peer circle (Berndt, 1999). Berndt (2002) documented that high quality of friendship is
recommended that negative friendship qualities can play more salient role than do
positive attributes (La Greca & Harrison, 2005). Quality of friendship is linked with
Bullying not only comprises violent actions involving offenders and targets, it is
relatively a dynamic issue pertaining to the social relationships of youth (Swearer et al.,
the network of friends and interactions with best friend. The degree of support afforded
by these relationships is also important in this regard as lack of support is associated with
risk of being bullied easily (Cook, Williams, Guerra, & Kim, 2010; Espealage &
Swearer, 2003).
69
maltreatment at school. Rothen, Head, Klineberg, and Stansfeld (2011) found that social
support from friends was more valuable as a protective dynamic than social support from
the family. Moreover, there was evidence that a high level of support from friends could
protect the targeted youth from poor achievement at school. Social support from family,
peers and school was associated with reduced internalizing distress resulted from
victimization experience. Surprisingly, close friend’s support did not moderate the
relationship. Moreover, social support did not protected victims from externalizing
distress.
Bollmer et al. (2005) investigated the friends’ relationship quality of bullies and
victims. The study confirmed the previous findings that friendship quality played a
that children having a higher quality best friendship will involve in bullying behavior less
often than children with a lower quality best friendship. High quality friendship not only
protected children from being targets of bullying, it also prevented children from
revealed that bully-victims did not differ from uninvolved children with respect to
conflict resolution, closeness and security, yet help was not considered to be significant
in preventing victimization. It implies that not all the dimensions and features of
friendship quality serve as a guard against bullying; therefore investigators should look
70
for the dynamics. Research has provided evidence in this regard as victimization is
associated with low levels of support and companionship within friendships (Hodges,
Boivin, Vitaro, & Bukowski, 1999; Rigby, 2000). Moreover the interpersonal risk of
victimization in relation to friendship quality should take into account the aspect of
reciprocity to deal with the discrepant findings (Rodkin & Hodges, 2003).
and students’ academic performance. Recently, Espelage, Hong, Rao, and Low (2013)
reviewed both cross-sectional and longitudinal studies in order to identify patterns of this
important predictor of decreased school grades and lower class room engagement during
middle school years. Not only chronic victims, but those who experience short-term
bullying episodes also reported impaired academic performance (Juvonen, Wang, &
Schools representing better grade averages had lower rates of bullying incidence that
indicates the value of improving school climate (Hanson, Austin, & Zheng, 2010).
and academic performance, Nakamoto and Schwartz (2010) pointed out some
discrepancies as well. Woods and Wolke (2004) demonstrated that both of the variables
are not associated to each other. Espelage et al. (2013) concluded that internalizing
behavior patterns particularly depressive tendencies used to play mediating role in this
relationship. Quality of friendship, perceived social support and school related issues
71
have also been regarded as moderating variables in relation to peer victimization and
academic performance (Erath et al., 2008; Schwartz, Gorman, Dodge, Pettit, & Bates,
2008; Simons-Morton, 2003; Wang, Iannotti, & Luk, 2011). Mechanism that account for
the link between victimization and academic performance may have practical
researchers (Scharphorn & Laura, 2013; Nelson & DeBacker, 2008). Urberg,
(e.g. smoking and alcohol use) friendship in groups. They found more insightful findings
about academic achievement that was associated with close friendship. A study by
Altermatt and Pomerantz (2005) also supported the notion that friendship plays a positive
role in the context of middle school students’ academic achievement. Other studies found
that children who connect with friends who reject school are more apt to achieve poorly
in academics (Veronneau, Vitaro, Pedersen, & Tremblay, 2008). Juvonen and Wentzel
(1996) looked into the social relationship processes counting interpersonal interactions to
observe how positive and negative qualities of friendship and the overall magnitude of
bullying behaviors. Fostering positive peer interaction can have an impact on school
grades.
72
well documented literature (Goetz, Frenzel, Pekrun, & Hall, 2005; Kyllonen, Lipnevich,
Burrus, & Roberts, 2010).In this vein, most of the studies focused on the relationship
Pakistani studies have also confirmed the existence of positive relationship between
emotional intelligence and academic achievement (Adnan, Chaudhry, & Malik, 2012;
information can improve cognitive and social functioning (Mayer & Salovey, 1997;
Salovey & Mayer, 1990).Yet this ability largely depends on accurate identification of
cross sectional sample of children, adolescents, and adult age groups for a Pakistani
sample (Khawar, Malik, Maqsood, Tallat, & Habib, 2014). Score on Ravens’ matrices is
2012).
intelligence. As the both of the latter variables have been recognized as a precursor to
could be one of the numerous potential pathways by which affective component persuade
academic achievement and may provide a new avenue for further investigations.
73
bullying is available beyond 5 to 6 years. Ahmer et al. (2008) raised the issue in medical
college students while finding a high rate of 63% students having faced the problem
during college years. Verbal assaults were the common form of bullying. Yet the study
The first prominent effort about school bullying was made by Hanif (2008) by
carrying out a cross-cultural comparison between Pakistani and English adolescents (14-
16years) about their perceptions and attitudes towards bullying and school social climate.
from both countries. Results suggested that culture played no significant role in
association among school climate, attitudes and perception towards bullying. School
social climate served a significant precursor to pupils’ perception and attitudes towards
bullying among samples of both countries. However English students perceived bullying
as more crucial issue and reported higher rates as compared to Pakistani students. In a
later investigation; Hanif, Nadeem, and Tariq (2011) found that attitudes of teachers,
Surprisingly the sympathy was not less pronounced for bullies. Such pr-bully attitude
Ahmed, Hussain, Ahmed, Ahmed, and Tabassum (2012) studied the effect of
estimates were obtained through teacher reports, according to which 74% teachers
acknowledged bullying as an issue in their school while remaining denied the fact. Nearly
similar ratios were obtained about students’ reporting of bullying experience to their
teachers. Most of the teachers (85%) favored anti bullying intervention. Verbal bullying
Association has revealed some facts regarding school bullying in Pakistan. The findings
health that had been in 2009. A large sample of 4647 students of grades 8 to 10 reported
their victimization experience during past month. More than 40 percent students were
found involved in bullying and victimization and boys outnumbered girls. Victims were
disturbances.
Only one study to date has empirically investigated the issue through self-report
(Illinois Bullying Scale) using behavioral descriptions about bullying, victimization and
fighting (Shujja, Atta, & Shujjat, 2014). The sample consisted of 836 students of 6th
grade from both public and private schools. Using a time frame of one month, almost
equal number of children reported bullying other or being bullied (23.2% & 24.1%)
respectively. Almost twenty percent students engaged into fighting with each other. All
the three aspects of bullying were more prevalent in public school. Interestingly
socioeconomic status did not predict involvement in bullying, yet it accounted for more
75
aggression among students. Gender differences were not prominent in prevalence yet
mean differences found that boys scored higher on bullying others, experiencing
The dearth of investigation in the vital aspect of bullying has left gaps that may
account for prevailing violence in Pakistani society. The researcher therefore decided to
victimization among the most vulnerable age groups (9-12 years old) as suggested by
existing literature (Due et al., 2005).This particular age group mostly studies in
elementary school (4th and 5th grades) and also includes beginners of middle school
(6thgraders). During the transition from elementary school to middle school, students
come into a new set up where the nature of peer groups are changing, academic work
becomes more rigorous, and biological changes start occurring within the individual
certain social and emotional deficits which might be explained by underlying cognitive
mechanism for example mood states (Lemerise & Arsenio, 2001). Mood state seems to
be a plausible contextual factor that can influence face perception. How mood state
mechanism has also produced inconsistent findings regarding mood congruency and
affect infusion paradigm (Schmid & Mast, 2010; Chepenik et al., 2007). Taking the
interactions into consideration, the current study primarily intended to explore how mood
Putting the emotion recognition into the perspective of SIP, we can speculate that
recognizing emotions in a peer interaction represents the first two stages of the model:
expressions may lead to misleading evaluation and biased attribution of a peer’s intend
which makes the situation and goals ambiguous and unclear. It further results in dys-
regulated arousal and angry appraisal, following which a child makes a faulty decision
and opts for an inappropriate action (Camodeca et al., 2003). Predisposed mood of
represents the child’s mental state at that time (see Figure 1.1).
skills and understanding, some investigators have emphasized that some bullies would
need good social cognition in order to perpetrate negative actions against victims and to
maintain a dominant status within the group. Thus, like theory of mind skills, emotion
bullies, victims, bully-victims and uninvolved students, and that too reflects
assuming better emotion recognition ability among bullies as compared to victims and
other children. However we have followed SIP model in explain the affective processes
parsimonious steps which could be easily assessed. Moreover explanation by Sutton and
mechanisms of bullying and victimization that was taken into account in present study.
with highly positive features also facilitates emotional competencies (Bollmer, Mirchi,
Harris, & Maras, 2005). Therefore we intended to investigate how bully/victim groups
would differ in their positive and negative aspects of friendship quality. Additionally, we
aimed to establish the possible link between friendship quality and emotion recognition
abilities of bully/victim groups. This particular aspect has not been studied by
researchers.
no significant data is available in Pakistan. For that reason, we also opted to study the
friendship quality and emotion recognition ability were also studied in relation to
behaviors. These facets of bullying have never been studied in Pakistan; where violence
78
and intolerance among youth is rapidly increasing. No large scale anti-bullying efforts
have been made so far neither any statistics is available officially. Present research would
provide the academicians and mental health professionals with a baseline to proceed with
such strategies.
79
Figure 1.1
General Problem
solving
Strategies
Formulate Response
Social Goals Decision
Child’s Mental
State
Social experiences
Social
expectancies
Knowledge of
social rules
Emotionality
Enacting the
Interpret Social Mood Response
Cues
Peer Evaluation
Emotion
and Response
Recognition
Encode Social
Cues
80
Figure 1.2
• Mood States
Affect = Bully
Mood +
Emotions
• Emotion
Recognition Involvement
in Academic
Performance
Bullying
• Quality of Victim
Peer Friendship
Relationships
81
Research Plan
The current research was carried out to determine the effect of mood on emotion
recognition and quality of friendship among school bullies and victims (see Figure 1.2).
To fulfill this purpose four studies were conducted. Each study served distinct objectives,
used separate samples for investigation, and followed diverse techniques and methods.
The first study was conducted to accomplish one of the prerequisites for the main
study. Since no indigenous measure was available, both bullying and quality of friendship
Qualities Scale (FQS) respectively. So it was mandatory to translate and validate these
instruments for use with school students. Therefore this study was carried out to translate
the measures into Urdu and determine their suitability for further use. This aim was
achieved in three steps for each of the measures; first a rigorous and standardized
procedure of translation was completed and secondly initial psychometric properties were
determined for OBVQ and FQS. Lastly the construct validity was confirmed for larger
samples. These steps are explained in three phases for both Olweus Bully/Victim
Questionnaire and the Friendship Qualities Scale. All the findings are summarized in the
bilingual sample of 36 students (Sample 1).The translated version of OBVQ was then
evaluated for initial psychometric properties and suitability for Pakistani sample. In this
regard, Exploratory Factor Analysis (EFA) was conducted and reliability estimates were
obtained along with providing initial evidence for convergent and discriminant validity of
the Urdu version on a sample of 122 students (Sample 2). Lastly, data for the main
studies (2, 3 & 4) were collected. Initial psychometric properties were further validated
using Confirmatory Factor Analysis (CFA) on the larger sample of 817 students (Sample
3). The results of self-report (OBVQ) were also cross validated with teacher nominations.
Phase 1: This measure was also translated and adapted into Urdu.
students (sample 1). The translated version of FQS was analyzed for initial psychometric
properties and suitability for Pakistani sample using EFA with 151 participants (Sample
2). Reliability estimates were also obtained and initial evidence for convergent and
discriminant validity of the Urdu version was provided. At the end FQS was further
Study 2 was carried out to obtain one of the key objectives of the present research
that was prevalence estimates of bullying and victimization. The results are based on the
83
data obtained in the 3rd phase of study 1(Section A). Findings of this study are based on
bullying and victimization across gender and three grades (4 th, 5th and 6th) in both public
and private sector schools of Lahore. Results for types, severity, duration, location and
Section B. This section included the details about demographic and academic
variables obtained for bully/victim groups. Monthly income, family size, parent’s
education, average marks in last two exams, average attendance percentages of last two
among school bullies and victims. This was not possible without an effective and
indigenously appropriate technique and procedure for inducing sad, happy and neutral
mood states in children. Therefore we first developed and evaluated mood induction
procedures for children and then proceeded with the main study. Hence the findings are
this phase was developing and standardizing mood induction technique for children. This
area had never been taken into account in Pakistan, so there was a need to generate an age
84
empirically investigated in this regard for example instrumental music, video song,
music. These four techniques were separately evaluated for happy, sad and neutral
randomization procedures.
Victims. This phase employed experimental procedures on a sub set of the sample drawn
from schools during study 1. It dealt with the randomization of sample for mood
conditions (happy, sad, and neutral) and mood priming using MIP from the first phase of
participants was evaluated pre and post induction of positive, negative or neutral mood.
Thus a mixed factorial design was followed for the current study. Emotion recognition
ability of the bullies, victims, bully-victims and uninvolved students was examined.
was examined.
The sample for this study (N = 672) was the same that was evaluated for CFA in
the third phase of section B in first study. It was also a subset of the larger sample of
bully/victims identified during study 1(N = 817). Besides investigating gender and grade
differences in students’ friendship quality, the study established the relationship between
friendship quality and bullying. In addition, relevance of friendship quality and bullying
with academic performance of students and demographic variables was also examined. It
85
In the end, all the findings were integrated. The sample of bully/victim groups
having scores on emotion recognition task, quality of friendship and academic variables
were analyzed for possible relationships, effects and interactions. Meaningful conclusions
were drawn.
The above mentioned research plan and organization of the dissertation is illustrated
below in figure1.3.
86
Figure 1.3
Chapter-II
MEASURES
rigorous exercise. Numerous guidelines and procedures have been offered in this regard.
measures acknowledged that there is a lot of diversity in the techniques and their worth
(Maneesriwongul & Dixon, 2004). A consensus has been found in the initial steps
followed by a committee review and pilot study of the both versions or at least the target
language version of the measure on a small sample (Beaton, Bombardier, Guillemin &
Ferraz, 2000). After making necessary amendments, the final translated version is then
assessed for its psychometric properties with a larger sample, e.g. at least 5 participants
against each item on a scale in exploratory factor analysis (Nunnally & Bernstein, 1994).
Other steps for example cognitive debriefing or focus group can be taken into account
before conducting a pilot study if these experts cannot achieve consensus or suggest
cultural consideration (Beaton, et al., 2000; Chavez & Canino, 2005; Hilton &
Skrutkowsky, 2002; Sousa & Rojjanasrirat, 2010). Considering all these guidelines, the
present study incorporates a five step procedure during first phase of the translation
process to achieve the cultural equivalence for Olweus Bully Victim Questionnaire and
Thus the first phase of the translation process included forward translation and
synthesis, back translation and synthesis, cognitive debriefing, committee review and
preliminary testing with bilingual sample. The second phase comprised of establishing
analysis, reliability coefficients, construct validity and the descriptive statistics. It was
1. To translate and adapt the Olweus Bully Victim Questionnaire (OBVQ) and Friendship
2. To explore the structure of OBVQ and FQS Urdu versions using EFA in order to check
3. To confirm the factor structure obtained from EFA on a larger sample (of main study)
3. To determine the psychometric properties of the Urdu versions of OBVQ and FQS.
Scale provided the measures on researcher’s personal request and allowed to translate and
adapt them for Pakistani sample. The translation process and evaluation of psychometric
properties of both the scales was carried out independently at different time periods with
different samples. The process and findings are presented in two sections.
type format by selecting from a response set having two to six possible options. OBVQ
begins with asking students’ like/dislike about their school, their gender and the number
of friends they have. The measure then provides a definition of bullying as purposive,
repeated and persistent, hurtful behavior that takes place within the context of an
between friendly, ordinary fights and the act of bullying. It provides the reference period
that is considered to represent a normal memory unit for the students (e.g. ‘‘in the past
couple of months,’ approximately representing the time period from the beginning of
school after the summer or winter vacation up to the day of the questionnaire
administration). This duration leads to more accurate recall of bullying experiences and
has been recommended by the author (Olweus, 1993), Most of the questions have clear
bullied) and bullying (bullying others) along with 9 types each (assessing verbal,
physical, relational, racial and sexual and cyber bullying) can be answered on a 5 point
scale with ‘‘I haven’t been bullied/bullied other students at school in the past couple of
months,’’ coded as 1 and “several times a week” coded as 5. These global questions
determine the participant’s role in bullying behavior. Other questions include information
.90) and was designed to be administered in a group setting (Solberg & Olweus, 2003).
Moreover, it can produce clear factors for items assessing bullying and victimization
(Bendixen & Olweus, 1999; Woods & Wolke, 2004). It has also demonstrated good
construct validity (Olweus, 2001). The OBQV is significantly associated with measures
2003).
The authors’ permission to translate the questionnaire into Urdu was sought. The
translation process followed the five steps procedure including a qualitative inquiry of the
PhD in child and educational psychology and having fluency in Urdu and English
languages were approached and instructed to translate the English OBVQ into Urdu by
emphasizing the intentions of original version and clarity of concepts during Urdu
translation. Both translations were evaluated by a committee of three experts in the field
note that no consensus was made for translating the word bullying as both translators
Step 1. Sixty students (30 boys and 30 girls) between 9-12 years (M = 10.63, SD =
1.05) studying in three private tuition centers (after school evening study centers) were
included in this investigation. Researcher enlisted all the behaviors and actions
corresponding bullying according to the Olweus Questionnaire and asked the participants
Table 2.1
Gender Grade
Badmashi 16 4 20 4 5 11 20
Gunda gardi 14 4 18 2 6 10 18
Sataana 6 11 17 6 7 4 17
Robe dalna 5 7 12 2 6 4 12
Dhouns jamana 3 6 9 0 2 7 9
Larae jhagra 1 5 6 4 2 0 6
Zulm karna 3 3 6 3 2 1 6
Fasaad 2 3 5 0 2 3 5
Dehshat gardi 3 0 3 0 1 2 3
A total of 10 terms were identified. Six of these terms having greater frequencies
were presented to the expert committee for discussion. Two of the identified terms
92
“badmashi and gunda gardi (hooligan)” though reported frequently were excluded due to
ethical reasons. Experts considered them inappropriate and sensitive to be used with
young children. Moreover, these terms were more likely to be reported by boys than girls.
Remaining four terms (Bohat tang karna, Sataana, Robe dalna and Dhouns jamana) were
above mentioned procedure, recall method was employed (Smith & Monks, 2008).
Another sample of 20 students (10 boys and 10 girls) (M = 10.45, SD = 1.09) was drawn
from a private tuition centre. They were presented with the list of 4 identified terms
representing bullying, and were asked to express them each in detail by recalling their
experiences at school. They were instructed to describe behaviors relevant to the listed
terms one by one. They were encouraged to report whatever comes to their mind after
reading each term. Table 2.2 shows the frequency of different types of behaviors,
categorized according to the types and nature of harmful actions described within each of
Step3. The results were analyzed by the expert committee and two most
comprehensive and suitable words were chosen as an alternate term for bullying and were
incorporated in the Urdu version of OBVQ (i.e. “Had say ziyada tang karna” and
“Dhouns jamana”). Both of these words were incorporated to the synthesized Urdu
version.
93
Table 2.2
cultural background and experience independently translated the Urdu version of the
questionnaire back into English. The expert committee again reviewed both translations
and came up with a converged version of reverse translation that was further compared to
the original version and was found identical and acceptable in terms of semantic
equivalence.
private and public sector schools studying in an academy in the evening was drawn and
the translated version was administered to them. They were encouraged to reflect upon
ach item with practical suggestions. Their insight, comprehension and understanding of
the measure were thought to be important for achieving the pragmatic and semantic
equivalence of both source and target language. The review committee again modified a
few statements and words to obtain parsimony by using language that could make the
items clearer to the children for example the term sexual typically translated as jinsi was
replaced with gandi baten (in item numbers 12 and 32 of OBVQ) as a result of cognitive
debriefing phase. Moreover, part of OBVQ item 18 which requires to mention the place
where bullying occurred was adapted culturally. Thus a pre-final Urdu version was
produced.
study was to empirically evaluate the OBVQ Urdu version for a sample of students
studying in 4th, 5th and 6th grades. Another aim was to determine the concurrent validity
of the OBVQ-Urdu version and to serve this purpose; responses from the original English
95
OBVQ version were compared with those from the translated Urdu version using
We explored the linguistic and content equivalence of the OBVQ and examined
the psychometric properties of the OBVQ Urdu version. Structure of OBVQ was
explored using exploratory factor analysis (EFA) that was later confirmed with CFA.
Three samples were drawn separately from schools to achieve these aims.
Linguistic equivalence.
Sample I. Three English medium schools were contacted and informed about the
purpose of study for obtaining a sample of bilingual students. Only one school agreed to
participate. Finally, a sample of 36 students, 18 girls and 18 boys studying in 4 th, 5th and
6th grades was drawn with the age ranging between 9 to 12 years (M = 10.53, SD = 1.1).
Measures. Both the original Olweus Bully Victim Questionnaire in English and
Procedure. After the school authority allowed for data collection, researcher
contacted the class teachers to provide a list of 45 students (15 students from each class)
whom they considered to be fluent in Urdu and English reading, writing and expressing.
They identified 36 students who were gathered to a separate room. The participants were
96
also informed about the confidentiality of data and their right to withdraw during the
study, despite their previous consent. The participants first completed the Urdu version of
OBVQ during 40 minutes session of regular school hours. The definition of bullying was
read aloud to them and they were asked to complete the instrument by keeping this
definition in mind. One week later, the same participants completed the English version
of the instrument. The items in both versions were presented in a different order.
study includes exploratory factor analysis and the initial psychometric evaluation of
Sample II. The sample for the empirical evaluation of psychometric properties of
the measures included 122 students (Mage = 11.13, SD = 1.1) studying in 4th, 5th and 6th
grades. Girls (n = 70) and boys (n = 52) were drawn from two private schools of Lahore.
Their mothers were also approached with the help of school administration, who rated
and two set of items measuring victimization (item 4-12a) and bullying (24-32a) were
children and adolescents. The present study assessed the Anxious/Depressed Domain and
Aggressive Behavior Domain for establishing the construct validity of the OBVQ. The
measure has already been translated into Urdu. The translated version has been used in
several investigations which have reported adequate psychometric properties (Anjum &
Malik, 2010).
Procedure. After obtaining formal permission from the principals of the schools,
the consent letters were sent to the parents of students seeking permission for assessing
their children. Mothers were also requested to fill the enclosed questionnaire. It included
Out of 140 requests, 18 were turned down thus the final sample constituted of 122
students and their mothers. Olweus Bully/Victim Questionnaire was administered to the
students in their regular class rooms. The seating plan was roll number wise that has been
arranged by the class teacher as per researcher’s instruction. After instructing the students
about filling the questionnaire, they were asked to answer the first two questions. The
definition included in the questionnaire was read aloud to them. They were asked to
mention any difficulty while answering the questions. It took only 15 minutes to
the questions. The questionnaire was filled anonymously and the information provided by
mothers was matched to the student data by code numbers mentioned on both of the
measures and also against the list of students provided by class teacher.
98
Confirmatory Factor Analysis and reliability estimates. This part includes the
Confirmatory Factor Analysis of Olweus Bully Victim Questionnaire (OBVQ). CFA was
computed in order to examine as how well the present data for main study (obtained from
sample III) fitted to the 2 factor structure of the questionnaire. It is a common practice in
CFA (Prooijen & Kloot, 2001). This step deemed important in current study due to the
cultural diversity of populations. Cronbach’s alpha wase also computed to determine the
reliability of the instrument. In this section, gender and grade differences were also
and uninvolved students were identified with the help of global items of Olweus
Bully/Victims Questionnaire (Urdu Version) and teacher nominations were also obtained
Sample III. The sample consisted of 817 students studying in 4th, 5th and 6th
grades of both public and private sector schools of Lahore. Girls (n = 440) and boys (n =
377) with an age range between 9-14 years (M = 10.89, SD = 1.12) completed the
Revised Olweus Bully Victim Questionnaire-Urdu version during school hours. Total 12
schools participated in this study and the 6-16 students were randomly selected from the
classes depending on the class size. Exclusion criteria included failure to complete the
majority of the questionnaires, and not indicating age or grade. Overall data of 817
Class teachers of these students were later asked to nominate them in one of the
four bullying roles: bully, victim, bully/victim and uninvolved. A total of 76 teachers
99
participated in the nomination procedure (M = 29.79, SD = 4.8). All the teachers had been
A full questionnaire was administered to the participant, yet items measuring different
forms of traditional bullying (25-32) and victimization (5-12) were evaluated using CFA.
Global items were excluded being separate indicators and the criterion for involvement in
bullying and victimization (Solberg & Olweus, 2003). Cyber bullying and victimization
related items (32a and 12a) were also excluded from the proposed model.
using definition of bullying from Revised Olweus Bully Victim Questionnaire. Teachers
were instructed to read the definition of bullying carefully followed by the description of
all the bullying roles: bully, victim, bully/victim and uninvolved. They were then
presented with the list of randomly selected students from their classes and were asked to
assign them to one of the four bullying roles based on their past three months observation
authority. The teachers were instructed not to share or discuss this information in the
class or with anyone else in the school. The measure was pilot tested with 5 teachers who
information within 10 minutes. The nomination method was preferred over rating
Procedure. A total of 30 (15 public and 15 private sector) schools from different
area of Lahore city were invited to participate in the study. Finally 12 schools; 6 public
sector schools and 6 private sector schools provided their consent for participation in the
current investigation. Detailed information was provided to the schools, however they
did not allow the researcher to visit regular classes or administer the questionnaire to the
whole class considering it a disturbance in school routine; rather 10-16 students were
randomly selected from each class. Permission was sought from the parents of these
Olweus Bully Victim Questionnaire follows the whole class approach, so despite
many restrictions by most of the schools, nearly 1000 detailed informed consents were
sent to the parents with the help of class teachers. They were first given one week to
return the consents that was later extended to one more week. However, unfortunately the
turnover was very poor (28%) especially for public sector schools (19%). The reason
could be low level of parental education or the students/parents misplaced the form and
verbally reported their consent. In the view of this experience passive consent was
participation in research in writing or do not send the form back. Under passive consent
procedure, parents are requested to report in writing only if they do not want to allow the
child to participate (Baker, Yardley, & McCaul 2001; Ellickson & Dawson, 1989). This
public sector schools in the present study. It contained all the necessary information about
the study and briefly described the procedures. Demographic sheet was also attached to
101
it. Parents were given two weeks’ time to report their concerns. Only 102 refusals were
received during two weeks, leaving 898 potential participants for the current study.
Following the exclusion criteria (mentioned above), data of 817 participants was used for
Each school provided a set time for students to take the survey during school.
school hours over a 45 minutes session. Students were encouraged to reflect upon their
school life during the last two to three months. Researcher read a set of standardized
instructions to the respondents about filling the questionnaire. They were also provided
with a definition of bullying for similar understanding of bullying phenomenon. All the
items of the questionnaire were read aloud for students in 4th and 5th grades for their better
comprehension. The next day, teachers were also provided with the list of students drawn
from their classes and they were requested to nominate them into one of the four bullying
roles. They first read the definition of bullying (based on OBVQ) and then the general
criteria for involvement and un-involvement in bullying. For this purpose four status
types (pure bullies, pure victims, bully-victims and the completely uninvolved) were
defined in the light of Olweus criteria of two to three times a month. Data from teachers
was mostly collected during break time. Bully-victim status as a result of self-report was
Results
obtained from bilingual sample (Sample I). Second phase includes the results of
exploratory factor analysis, convergent and discriminant validity of OBVQ, the reliability
coefficients and mean scores for two set of items representing bullying and victimization
(Sample II). Later we presented findings of CFA and further support to internal
Table 2.3
Age
Table 2.3 shows the frequency distribution of samples used to establish the
linguistic and semantic equivalence of OBVQ and the initial psychometric properties of
Urdu version. The bilingual sample of school students (sample I; N = 36) was fairly
distributed across gender as 50% were boys and 50% were girls. Majority of the students
belonged to 5th and 6th grades. It consisted of 42.6% girls and 57.4% boys with 34.4%
104
students studying in 4th grade, 38.5% students in 5th grade and 27% were studying in
grade six. Differences in age of both samples are due to the type of schools from which
they were drawn. Sample I was drawn exclusively from a private school, whereas sample
II was recruited from both public and private schools. Sample III consisted of 53.9% girls
and 46.1% boys between the ages of 9-14.Greater proportion of the sample (48.8%)
studied in 6th grade while 26.1% studied in 5th grade and 25.1% used to study in grade 4.
First of all inter-item correlation was computed between English and Urdu
versions of the OBVQ. There was one week gap between the administrations of both
Table 2.4
Items r Items r
Table 2.4 shows the inter-item correlations between the English and translated
Urdu version of Olweus Bully Victim Questionnaire. Results indicate that some of the
items (7, 10, 12a, 24, 29, 35 & 38) showed very high correlations (r > .90). Strength of
106
correlation was still quite strong (r > .80) for most of the items (3, 4, 5, 6, 8, 9, 11, 15, 16,
17, 20, 21, 25, 26, 27, 28, 31, 32, 32a, 34, 36, 37 & 39) whereas a few items (1, 12, 13,
14, 19, 22, 23, & 30) were moderately correlated (r > .70). Relationship strength of only
two items (18, 33) was slightly below .70, yet all the correlations were highly significant
(p < .001). Item number 2 showed perfect correlation as it indicates the gender. To
summarize, we can say that all the items in the OBVQ Urdu version were comparable to
Table 2.5
Scale No of items K α
Results in Table 2.5 show that both English and Urdu versions of OBVQ yielded
There has been lot of argument about factor analyzing techniques when the
analysis (CFA) is preferred when the population of interest is the same, yet one should
not ignore the cultural diversity of population when a translated measure is administered
in a different country. Thompson (2004) has suggested a three step procedure that
107
includes meaningfully reducing the number of items using Principal Component Analysis
(PCA) first of all and then extracting the latent variables using true Exploratory Factor
Analysis (EFA). Lastly he recommended using CFA to verify the existing factor
structure. In this section, we opted to use EFA that was followed by a CFA computed for
factor structure have also been contentious (Pett, Lackey, & Sullivan, 2003). Researchers
have suggested that factor analysis is preferable to principal components analysis which
only reduces the data set (Floyd & Widaman, 1995; MacCallum & Tucker, 1991).
Common factor models intend to understand the underlying latent unobserved) variables
that explain relationships among the variables measured. Rotation is also considered
important to simplify and clarify the data structure (Costello & Osborne, 2005).
Orthogonal methods (varimax, quartimax, and equamax) are most widely used by
researchers to avoid complexity, yet these methods do not address the correlation among
factors that are best explained by oblique methods (direct oblimin, quartimin, and
promax) (Fabrigar, Wegener, MacCallum, & Strahan 1999). The present study was
conducted to see if the set of items in OBVQ measuring bullying and victimization
represent two distinct behavior patterns that could further distinguish the students into
different bullying roles. Prinicipal Component Analysis (PCA) with varimax method of
Sample size is another critical issue in EFA. Fabrigar et al. (1999) found that a
most of the studies (75.7%) reported variable to-factor ratios of at least 4:1. To check the
108
were computed.
Table 2.6
Kaiser-Meyer-Olkin Test of Sampling Adequacy and Bartlett Test of Sphericity for OBVQ
KMO was found to be .86 that exceeds the minimum value of .50 suggested by
(Field, 2005) and Bartlett’s Test of Sphericity yielded highly significant findings (χ2
(190) = 1733.3, p < .001), which showed the appropriateness of the data for factor
analysis.
Since SPSS has limited options for factor retaining methods, we relied on
typically retaining factors with Eigen values greater than 1.0 (Kaiser, 1956). Eigen values
not only indicate the significance of specific factor but also demonstrate the amount of
variance in a set of items accounted for by that particular factor. Scree plots were also
obtained to make the best choice in this regard. It is also important to select the
significance level for loadings. Tabachnick and Fidell (2001) suggested .32 as a good
criterion for the minimum loading of an item, which is equal to about 10% overlapping
variance with the other items in that factor. Here we chose minimum of .35 as loading
standard for an item that is recommended by Stevens (as cited in Field, 2005).
109
Table 2.7
Summary of Principal Component Analysis with Varimax Rotation for OBVQ Urdu
Version (N = 122)
Table 2.7 shows the factor loadings of 20 items from OBVQ Urdu version. These
items measure the bullying and victimization status of the students by asking about
general and specific behaviors that are often linked to bullying. The results showed high
communalities for all the items. A principal component analysis with varimax rotation
resulted in factor solution that converged in 3 iterations and high ladings ranging from
.68 to .84. The three factor solution showed Eigen values greater than 1 and the extraction
was confirmed by scree plots. The items indicating the global question about being
bullied (item 4) and different types of traditional bullying experienced (5-12) by students
made the first factor named Victimization (V). It had the highest Eigen value (7.50)
explaining the largest part (37.53%) of the total variance. Similarly, set of items asking
general question about bullying others (item 24) and related types of traditional
perpetration of bullying (items 25-32) resulted in the second factor that is called Bullying
(B). Factor II showed Eigen value of 4.41 which explained 22.09 % of variance. This was
an expected depiction except for two items that accounted for the third factor (item 32
and 12 a) having an Eigen value merely crossing one (1.50) and accounting for minimal
variance (7. 54%) of the total. Both of these items represent the cyber bullying and
victimization respectively. The third factor itself may not be meaningful having low
indices yet we decided to retain both its items as they measure an important aspect of
bullying and results can be subjected to smaller sample size. The total amount of variance
Scale item characteristics. This portion includes the reliability coefficients and
descriptive statistics of the extracted factors and total set of items. Additionally, it
111
presents means and standard deviations for each item, item total correlation and
Table 2.8
Item total correlations were computed to evaluate each item in order to confirm
whether all the items significantly measure the bullying phenomenon. Table 2.8 displays
112
that deletion of any individual item changes alpha values to range between .90 and .91.
Item to total scale correlations ranged between .41 to.74, except for two items measuring
cyber victimization (12a) and cyber bullying (32a) showing correlation of .01 and .08
respectively. However we decided to retain the items for final Urdu version, considering
the sample size constraints for current findings. Mean scores of victimization items (4-
Table 2.9
Descriptive Statistics and Reliability Estimates of the Olweus Bully Victim Questionnaire
Total 20 .911
The subscale of Bullying shows a mean of 21.05 and a standard deviation 10.78,
and the Victimization subscale shows a mean of 16.63 and a standard deviation of 8.56.
Both of the factors showed high estimates of reliability coefficient. Reliability of Cyber
bullying and victimization factor was also moderately high and mean scores were quite
consistency of the 20 set of items measuring bullying and victimization aspect of OBVQ
Table 2.10
Subscales 1 2 3
2. Victimization
- - .046
3. Cyber Bullying/Victimization - - -
Table 2.10 illustrates that bullying and victimization scales that represent
traditional nature of bullying experience are significantly correlated with each other (r =
.26, p < .01), yet the magnitude of this relationship is not so strong. Cyber bullying and
of the measure. To fulfill this purpose, two subscales of Child Behavior Checklist
Table 2.11
Convergent and Discriminant Validity of the Olweus Bully Victim Questionnaire (OBVQ)
CBCL Subscales
OBVQ and CBCL. Bullying subscale despite being correlated with anxious/depressed
dimension of CBCL exhibited stronger relationship with aggressive behavior (r = .78, p <
.001 vs. r =.41, p < .001). Similarly victimization was significantly related to anxious
depressed tendency (r = .49, p < .001) while showed no relationship with aggressiveness
(r = .15, p = ns). The nature of relationship provides some support to the convergent and
The process of cleaning and screening data included inconsistency checks and
missing responses has been considered an important step before conducting sophisticated
statistical analysis (Luck & Rubin 1987; Malhotra 1999). First section presents the
Confirmatory Factor Analysis on data for types of victimization and bullying. Items of
Olweus Bully Victim Questionnaire assessing global victimization and bullying (4 and
24) and subsequent set of items representing the types of victimization and bullying were
analyzed for missing values and descriptive statistics. Results are presented in Tables
Table 2.12
Missing values were found for only one variable (.1%) that was replaced with
mean. It is not unusual to obtain extreme values due to the nature of construct being
measured.
117
Table 2.13
Skewness Kurtosis
Table 2.13 indicates that data on the items of OBVQ was fairly normally
distributed except for items 12a and 32a that measure cyber victimization and bullying
respectively. Kline (1998; 2005; 2011) has suggested that skew value greater than 3.00
and kurtosis value greater than 8 can be problematic and may distort the results of
mentioned descriptive statistics, two items (12a and 32a) were not included in the
CFA
Structures (AMOS), Version 20.0. CFA is a structural equation modeling technique that
is used to evaluate the goodness of fit between a theoretical model and the sample data
(Hu & Bentler, 1999). Global items were also not included in the model as we intended
to check if the set of items assessing different types of victimization and bullying had
resulted in distinct patterns within this sample that could differentiate bullies and
victims.
view (see App. B), model was repeated with addition of relationship path between items
25 and 27 of bullying subscale. Table 2.14 presents the fit indices for both models.
119
Figure 2.1
Table 2.14
A poor fit was suggested by the chi square value being significant, χ2 (103) =
333.59, p < .05. Yet the CMIN/df value (CMIN/DF = 3.24) was found approximately
acceptable (Schumacker & Lomax, 2004). We also considered RMR for not being
influenced by chi-square or sample size (Kline, 2013) that was found within acceptable
for model fit (.045). Rest of the indices that have been considered for assessing the
degree of fit between the hypothesized two factor model and the sample resulted in a
Tucker Lewis Index and the Comparative Fit Index were found to be acceptable
having values .92 and .93 respectively for model 1 (Tucker & Lewis, 1973; Bentler,
1990; Hu & Bentler, 1999). Root Mean Square Error of Approximation yielded a score of
.05 that had been considered excellent (Brown & Cudeck, 1993). Considering
to 2.99 and increased the TLI and CFI up to one point each (.93 and .94 respectively).
PLCOSE value was .26 that adds to the support of good model fit. RMSEA decreased
(.04) and PCLOSE was improved (.56). Thus we obtained a well fitted model of OBVQ.
121
Table 2.15
Factor 1 Factor 2
Items (V) (B)
Factor 1: Victimization (V)
Factor loadings for victimization scale range from .54 to .71, and items on
bullying scale also showed high loadings ranging between .54 and .65. Moreover,
moderate correlation between (r = .50) the two factors, provided further evidence that
Table 2.16
Scales k α α α α α α
from .81 to .86 on the OBVQ-Urdu Version. Overall, the reliability estimate for 18 items
based on Cronbach’s Alpha, is also uniformly high across gender and grade level.
Table 2.17
Skewness Kurtosis
Victimization 817 8.00 40.00 14.2827 6.37951 1.025 .086 .402 .171
Bullying 816 8.00 40.00 12.6262 5.22958 1.250 .086 1.352 .171
7.40) more than reports of bullying (M = 15.02, SD = 6.22). Statistics for skewness and
123
kurtosis were approximately within acceptable range demonstrating that data were
Table 2.18
MANOVA showing Gender and Grade Differences for Total Bullying and Victimization
Source DVs SS df MS F p η2
differences in victimization scores, (F (1, 810) = 6.71, p < .01, η2 =.008). Boys scored
2.18, p = .14, η2 = .003) were observed across gender. MANOVA results provided
124
evidence for significant difference in victimization scores of three grade levels, (F (2,
810) = 9.51, p < .001, η2 = .023). Students in three grades (4th, 5th, and 6th) also
significantly differed in reporting perpetration of bullying, (F (2, 810) = 6.41, p < .01, η2
= .016). Interaction of gender and age did not significantly affected bullying, (F (2, 810)
= 1.27, p < .281, η2 = .003) and victimization scores, (F (2, 810) = 1.49, p < .226, η2 =
.004).
Figure 2.2
Figure 2.3
Table 2.19
95% CI
LSD post hoc comparisons among three grades revealed that students in 4th grade
scored significantly higher on the Bullying subscale while there were no significant
differences between students in 5th and 6th grades. Similar results were found on the
Victimization subscale.
127
Bully/Victim Groups
and bully-victims (n = 235). The rest were grouped together as uninvolved (n = 283). To
(1) The cases of the children that were involved in bullying others for two to
three times a month or more on global item of bullying (item 24) and at least one of the 8
subsequent forms of bullying, but had not been victims of bullying were classified as
bullies.
(2) The cases of the children that were bullied two to three times a month as
measured by global item of victimization (item 4) along with at least one of the 8 ways,
but had not been involved in bullying others, were categorized as victims.
(3) The cases of the children that had been bullying others and been bullied two to
three times a month or more on both global items and at least one of the 8 corresponding
check on the accuracy of the self-report. The estimated number of class bullies (and
128
victims) was compared to the total number of students who reported being bullies (and
victims).
Table 2.20
Urdu) has been correctly classified as evidenced by teacher nominations. Cohen’s κ was
conducted to determine the level of agreement between teacher nominations and self-
reported bullying classification for 817 students. Kappa showed high correspondence
received the same status in teacher nominations. High level of agreement (89.1%) was
the self-reported victims were accurately classified (87.7%) by teachers into the same
acknowledged their role in self-report (70.6%), some of them were identified as bully-
victims (13.9%), and victims (5.9%). A few (9.6%) reported them as uninvolved.
Table 2.21
Victimization
95% CI
TN OBVQ B SE Wald OR LL UL
Bully
victimization .31 .18 2.98 1.37 .96 1.95
bullying 2.33 .19 155.63* 10.28 7.13 14.82
Victim
victimization 2.38 .19 154.4* 10.75 7.39 15.63
bullying -.14 .21 .43 .87 .58 1.32
Bully -Victim
victimization 1.89 .19 102.61* 6.61 4.59 9.53
bullying 2.18 .19 125.09* 8.81 6.02 12.89
*p < .05.
130
Children who scored high on global bullying item (24) of OBVQ were more
global item (4) of OBVQ significantly predicted teacher nominations for victims. Teacher
nominated bully/victim group scored higher on both of the items respectively. The results
and bullying.
131
Discussion
The translation process of OBVQ produced fruitful outcomes while finding the
bullying and the word is repeatedly used in the questionnaire. While translating OBVQ
into Urdu, term Bullying raised some questions and contradictions. Therefore a brief
qualitative examination was preferred to answer this question. Smith and Monks (2008)
have also argued that finding a suitable equivalent for the word bullying while translating
two basic methods for such an investigation, recognition and recall. Recognition method
that is based on cartoon task has been widely used across countries for this purpose and
has provided appropriate alternates to bullying in Japan and Korea (e.g. Ijmi and Wang ta
respectively). Since it was not a main objective of the study, recall method was used for
identifying suitable alternate of bullying. This is one of the methods mentioned by Smith
bullying while inquiring students to report how often they bully others or are bullied
(Vaillancourt et al., 2008). Therefore this exploration was considered necessary before
finalizing the translated version of OBVQ. It brought about meaningful insight to the
phenomenon as the children provided the account themselves. The top listed terms
‘Badmashi’ and ‘Ghunda gardi’ were not retained due to ethical issues.
132
Students’ responses for the four terms ‘Had se zayada tang karna’, Robe jamana’,
‘Sataana’ and ‘Douns Jamana’ were evaluated. This process also resulted in a definition
of bullying provided by Pakistani school students which reflects some of the aspects
intimidation, verbal harm, harm to property, social and psychological harm, sexual harm,
power exhibition, and coercion were distinctly recognized as essential components of the
documented that majority of the pupils had been unable to identify ‘Imbalance of power’
as a component of bullying; rather ‘intent to hurt’ was acknowledged as the core feature
of bullying (Gordillo, 2011; Vaillancourt et al., 2008). Though present study was not
gender, age and grade-wise data of this part was not taken into account. However
findings draw somewhat similar conclusions that intention to harm was the main focus
described into physical, verbal, social, psychological and sexual types. Power and
characteristics of bullying was considered suitable yet faced the issue of complexity.
Students in grades four and five (mostly below 10 years) might not understand the word.
Therefore the final Urdu version incorporated both ‘Dhouns jamana’ and ‘Had say
zayada tang karna” (see App. A). It was further suggested to use Dhouns Jamana or
Ghunda Gardi for adolescent version of the questionnaire after investigating their
understanding of the phenomenon. The investigation was beneficial and provided suitable
133
within sentences.
Furthermore, discussions with two eminent experts also lend support to the
December, 24, 2014). Most recently, Smith et al. (in press) evaluated data obtained from
cartoon task (recognition method) from different parts of the word including Pakistan.
Similar terms (i.e. Ghunda Pan, Tang karna and Dhamkana) emerged from the study.
Further investigation with regard to the perception of the terms identified in this study
and the study conducted by Smith and colleagues in Pakistan could explain the
Psychometric evaluation.
between original and translated versions of OBVQ. It also provides the ground for
factor solution. Item for being bullied and bullying other were separately loaded which
demonstrates that the measure is suitable for identifying students involved in bullying. It
can clearly differentiate between bullies and victims in schools. However, items
measuring cyber bullying and victimization were loaded distinctively on a separate factor
indicating that the students who were involved in bullying others using cyber technology
had also been victim to it. Cyber bullying has gained attention of the researchers recently
and has been linked to traditional bullying (Hinduja, & Patchin, 2010; Ybarra & Mitchell,
134
2004). It is noteworthy that students in present sample had an average age of 11. 13 years
and the sample size were also small. Children at this age usually have access to mobile
phones and internet under adult supervision especially in the context of Pakistani culture.
Therefore, further inquiry with larger sample size is required to confirm the factor
The internal consistencies of the OBVQ total scores and for the subscales were
computed using Cronbach’s α coefficient. Results showed higher reliability estimates for
victimization and the total score (> .90), still good internal consistency for bullying
subscale (.80) and moderate reliability for items measuring cyber bullying and
victimization (.78). It also proved to be reliable across grades and gender. Previous
investigation with larger sample was required to confirm the findings. Moreover current
Urdu version was found suitable for elementary school children. Separate Urdu version
must be used using with adolescents or high school and college students after
Nearly all the item-total correlations were above .40 demonstrating that these
items should be retained for questionnaire. Only two items (12a and 32a) showed below
average correlations. However considering the distinct nature of cyber bullying (as it also
emerged as a separate factor) and minimal impact on overall internal consistency of the
measure (.90 to .91), both of the items were retained in the final Urdu version.
computing the relationship between bullying victimization and two important subscales
135
of behavior respectively. Both of these scales were significantly correlated with bullying
and victimization scales of OBVQ. Yet the nature of relationship was stronger between
Both of the global items (4 and 24) were used as indicators of victimization and bullying
respectively. A number of existing studies have supported this association. Solberg and
Olweus (2003) reported that students who scored higher on global item of victimization
showed significant inclination towards depression. The discriminant value of global item
assessing bullying was demonstrated by significant differences between bullies and non-
bullies on level of aggression and antisocial behavior. Other earlier and recent studies
have demonstrated consistent support for positive association between peer victimization
and internalizing symptoms such as depression and anxiety and also strong relationship
between bullying and externalizing behaviors (Boivin, Hymel, & Bukowski, 1995;
Reijntjes et al, 2010). The directionality of the relationship is however questionable and
Confirmatory Factory Analysis. The confirmatory factor analysis model for the
current sample yielded acceptable to excellent values on all the fit indices except for chi
square that was found significant. It is not uncommon to obtain a significant chi square
value for model with larger sample size that results in overstating the lack of fit (Furr,
2011; Leach et al. 2008); so the overall adequacy of the model should be interpreted with
caution by taking other fit indices into account after dismissal of chi square (Goffin,
2007; Jackson, Gillapsy Jr, & Stephenson, 2009; Markland, 2007). Model could be
136
interpreted as good fit because of acceptable values (CFI = .93 , TLI = .92, and RMSEA
= .05) and supporting parameter estimates that present high loading on both the factors;
yet CMIN/Df was slightly higher (.32) than accepted range (< 2 and in some cases up to
3) interpret our model as a good fit. Only one change pertaining to modification indices
(correlating the errors of items 25 and 27) markedly made the difference in CMIN/DF
(2.99) and also provided better fit values on CFI and TLI (.94 and .93 respectively). The
correlation suggested that those who bully other verbally were also likely to bully their
The gender and grade-wise reliability estimates of both bullying and victimization
scales were high. Reliability of the scales was also high for the total sample. These results
(Solberg & Olweus, 2003; Kyriakides, Kaloyirou, & Lindsay, 2006; Panayiotis et al.,
2010). Initial concurrent validity evidence has been provided following the procedure
described by Shaw, Dooley, Cross, Zubrick and Waters (2013) that is high correlation for
Gender and grade differences were observed for overall involvement in bullying
and victimization. MANOVA result suggested that boys and girls were equally involved
in bullying as perpetrators while boys reported more victimization than girls. Gender
reporting over-involvement of male compared to female students and others suggest less
pronounced gender differences. Gender differences are more likely to be associated with
different forms of bullying for example boys involve more in direct forms as compared to
girls who are subjected to indirect forms of bullying (Rueger & Jenkins, 2014; Selekman
137
& Vessey, 2004). Additionally, students in 4th grade were more prone towards bullying
others and being bullied compared to 5th and 6th graders. The results are in line with the
previous findings which reported bullying as more prevalent in lower grades (Salmivalli,
2002; Seals and Young, 2003). It is important to mention that only a few of studies have
used the composite scale scores on OBVQ to determine overall gender and age
Probably the extent to which a student might be bullied lies on a continuum, rather than
Researchers have argued about the accuracy of OBVQ cut off for identifying
bullying and victimization. The standard cut-off suggested by Solberg and Olweus (2003)
has been considered to be too strict for classifying bullies and victims (Lee & Cornell,
2009). On contrary, others have offered more stringent criteria to cater the element of
repetition in bullying definition (Scheithauer, Hayer, Petermann & Juger, 2006). Since it
was the first study using Olweus questionnaire in Pakistan, we followed standard Olweus
criteria of “2 to 3 times a month” on both global items along with involvement in at least
determine if the two methods identified the same students in different bullying roles. The
results demonstrated that the teachers detected almost 87% of respondents accurately into
bullying roles who reported on the OBVQ; however, they also identified almost 30% of
respondents as having been bullied, who actually did not report bullying on the BVQ.
138
Percent agreement across two methods for classification of victim, bully victim and
uninvolved group was high. There is more support for self-reported victimization than
bullying. Hence the social desirability can be the reason for low reported bullying by
we conducted a logistic regression using the two OBVS global items to predict teacher
nominations. Results showed that respondents endorsing the OBVQ item for being
bullied were over 10.8 times more likely to be designated as victims by teachers than
those who did not endorse this The same was true for bullies with an odd ratio of 10.2.
Students who responded above the cut off level on both global items were more likely to
be nominated as bully-victim by their teachers. Existing studies have found only low to
(Beran & Stewart, 2008); whereas, current study has indicated moderate to high
The study provides strong evidence to the psychometric properties of the Olweus
Bully Victim Questionnaire Urdu version. The questionnaire can be used in identifying
information obtained from OBVQ data provides guidelines to the researchers and school
Section B. Friendship Qualities Scale (FQS; Bukowski, Hoza & Boivin, 1994)
of children’s best friendship (Bukowski et al., 1994). Five factors of friendship include
mostly positive dimensions such as companionship, help, security, closeness; and one
negative aspect namely conflict. The participants rate the accuracy of each statement
about their friendship with the close nearest friend on a 5 point scale. The
Companionship subscale includes four statements that represent those activities which
friends do together. The Help subscale is consisted of 5 items and assesses how much the
dyad facilitates each other. The Security subscale comprises 5 items that address how
friends protect each other in different situations such as victimization. The Closeness
measured by 5 items. The conflict subscale identifies the level of disagreement between
friends.
The FQS is a widely used measure having sound psychometric properties with
reliabilities ranging from .71 to .86 (Bukowski et al., 1994) and .66 to .89 for the sub
scales (as cited in Demmings, 2009; Demir & Urberg, 2004). It has been translated and
adapted cross-culturally and studies have either confirmed the basic 5 factor structure
(Atik, Cok, Coban, Dogan & Karaman, 2014) or resulted in a two dimensional model
I contacted the authors who provided me with the scale and allowed to me use it for
translators (PhD scholar Psychology and PhD in Urdu), whose mother language was
Urdu, translated the English version into Urdu language. They were instructed to translate
the scale for children and adolescents while considering the conceptual similarity. They
generated two translated versions including words and sentences that cover both the
psychological meanings and the culturally spoken tone of language. Both the translations
similarities were identified and the format and instructions were first finalized.
ensure fidelity of meaning. The synthesized scale was then given to two more
independent bilingual translators for backward translation. Each of them was enrolled in
PhD in Psychology and English Linguistics respectively in United Kingdom and had not
compared the directions, sentences and answer format of the two back-translations with
the original Friendship Qualities Scale regarding layout, language, and grammatical
141
structure of the statements. Similarity in connotation, and relevance was sought for items
and variations were discussed to achieve the pre-final version of the scale.
3. Cognitive debriefing. The pre final Urdu version of FQS was pilot tested
for clarity with a sample of 10 students recruited from the target population i.e. 9-12
years old studying in 4th, 5th and 6th grades of public and private sector schools. The
participants were read each question and were asked to 'think aloud' as they consider
what their answer was. This procedure was used to provide additional support to the
conceptual and semantic equivalency of the FQS Urdu version. It helped in improving the
sentence structure for easy understanding by the target population for example item
numbers 14 and 20 were revised as a result of feedback during this procedure. They also
reported that item number 18 (‘If my friend or I do something that bothers the other one
of us we can make up easily’; ‘agar mera dost ya mein koe esa kam kren jo hum dono me
say kisi k liye bhe pareshani ka baais ho to hum usay mil kar thik kar letay hain’) was
difficult to understand due to its length. But when they read it twice, they figured it out
easily.
psychological assessment, two bilingual forward translators and the researcher evaluated
amendments and finalized the Urdu version of FQS. The word “Lunch” in item 14 had
been translated as “dopehar ka khana” but students preferred to use ‘lunch’ as being more
familiar in that context. So we kept as it was. Translation of the word “disagree” in item
20 “My friend and I disagree about many things” was also revised. Urdu word “Ikhtilaf”
142
was difficult for children to understand; therefore it was replaced with “Ittifaq-e-raey
5. Preliminary testing with bilingual sample. This rarely used step was
considered for the current translation as a pre-field test among students who were fluent
We explored the linguistic and content equivalence of the FQS and examined the
psychometric properties of the FQS Urdu version. Structure of FQS was explored using
exploratory factor analysis (EFA) that was later confirmed with CFA. Reliability
Linguistic equivalence.
Sample I. Thirty two students studying in 4th, 5th and 6th grades (Mage = 10.21, SD
= .97) participated in this try field test. The sample size was smaller than recommended
due to the reason of unavailability of the bilingual students as a result of strict policies by
quality English medium schools that do not allow data collection to the researcher.
Moreover the data were collected during summer vacations; therefore I contacted two
private English medium schools offering summer camp. Only one school agreed to
Mesures. English and Urdu versions of FQS were used in the present
investigation.
randomly selected 40 students and consent letters were distributed to them. Parents of 21
students provided their consent. Data were collected during school hours. The
13 students were approached in an evening tuition center. All of these children had been
studying in English medium schools and provided parental consent for participating the
research.
The FQS starts with naming the best friend as a prerequisite. Many students asked
if they could name more than one friend. They were instructed to think about their very
best friend for a while and then write down one name only. The instructions were read
aloud to them and they were encouraged to think about their best friend, whom they had
named on the first page, while answering the questions. The students completed the Urdu
version of the Friendship Qualities Scale in first session. One week later, the same
students were asked to complete the English version of the scale. All students who
participated in the Urdu version administration were present that day. Only two students
in school setting expressed difficulty in answering the English version and showed
unwillingness to continue. They were thanked and sent back to their classes. More than
40% information was missing for another student, who was excluded from the final
sample eventually. Rest of the students did not report any reading or comprehension
Items of the original FQS were presented in a mixed order to appear in a different
way from that of the items of the FQS Urdu version. Responses on both versions of the
measure were then evaluated to determine criterion equivalency. This step also
establishes the preliminary technical equivalence and confirms the conceptual, semantic,
and content validity of the translated version of an instrument before conducting full
Urdu version to the sample of main study, we evaluated the suitability of the scale by
exploring the factor structure on a sample of 151 students and also determined the initial
psychometric properties.
Sample II. One hundred and fifty one students (Mage = 10.79; SDage = 1.08)
including 70 boys and 81 girls participated in this study. They were drawn from 4 th, 5th
FQS Urdu version. The translated Urdu version of Friendship Qualities scale was
subscale of an indigenous measure of children’s social competence scale was used. The
Social Competence Scale for Children SCSC) is 40 item measure of social competence
for children in middle childhood or early adolescence. It includes 6 factors that have been
145
labeled as Self-control, Empathy and Helping Behavior, Social Skills and Obedience,
Antisocial Behavior, Assertiveness and Communications Skills. The subscales used in the
present study were ‘Empathy and Helping Behavior’ (12 items) and Self-
coefficients having Cronbach Alpha of .74 and .75 respectively (Malik & Shujja, 2011).
Procedure. The data were collected from two private academies during summer
vacations. Permission was sought from the academy staff and the parents of students. I
distributed consent letters to 200 students, of which 151 approvals were received. The
students were provided with both the scales in a single session. They were instructed to
complete the set, however, 16 students did not complete the social competence subscales.
considered essential for confirming the results of EFA obtained in the previous portion of
Sample III. Sample for this part of study was derived from the main study sample
of 817 students (mentioned in the previous part). It consisted of 672 students (Mage =
10.86; SDage = 1.08). There were 319 boys and 353 girls.
FQS rather than 23 item scale. Urdu version yielded adequate reliability estimates (α =
.77).
Procedure. The sample which was assessed for OBVQ in last segment was also
assessed for the quality of their friendship. Friendship Qualities Scale was administered
the next day of assessment of self-reported bullying as the school administration spared
the students only for an hour each day. Moreover, two of the schools could not be
approached for the second session with students. Therefore 684 students completed FQS
following the procedure described in previous segment. Twelve students left the scale
incomplete (more than 3 items were left unanswered). Therefore we evaluated a sample
Results
Later phase illustrates the results of exploratory factor analysis which ensured the
dimensionality and internal consistency of the adapted version of FQS Urdu version
along with evidence for convergent and divergent validity (Sample II). Results of
convergent and divergent validity of FQS Urdu version have also been included in this
section. Lastly Results of confirmatory factor analysis are presented (Sample III).
Table 2.22
Table 2.22 shows the frequency distribution of samples used to establish the
linguistic and semantic equivalence of FQS and the initial psychometric properties of
Urdu version. The bilingual sample of school students (Sample I; N = 32) comprised of
53.1 % girls and 46.9% boys. Majority of the students belonged to 5th (43.7%) and 6th
(37.5%) grades. Most of the students (65.5%) were 9 to 10 years old. The age of students
in sample II (N = 122) ranged from 9-14 years. It consisted of 53.6% girls and 46.4%
boys with 26.5% students studying in 4th grade, 33.8% students in 5th grade and 39.7%
were studying in grade six. Differences in age of both samples are due to the type of
schools from which they were drawn. Sample I was drawn exclusively from a private
school; whereas sample II and II were recruited from both public and private schools.
Inter-item correlations were computed between English and Urdu versions of the
OBVQ. There was one week gap between administrations of both the versions.
149
Table 2.23
Items r Items r
Table 2.23 shows the inter-item correlations between the English and translated
Urdu version of Friendship Qualities Scale. Results indicate two items (3 &12) showed
very strong correlations (r > .90). Most of the items (4, 7, 9, 10, 15, 16, 17, 21 & 22)
demonstrated high correlations (r > .80), whereas others (1, 2, 5, 6, 8, 11, 13, 14, 19 &
22) were moderately correlated (r > .70). Relationship strength of only one item (20) was
below .70, yet all the correlations were highly significant (p < .001). To sum up, we can
150
say that all the items in the FQS Urdu version were comparable to those in English
Table 2.24
Reliability Coefficients for English and Urdu Versions of FQS (Sample I; N = 32).
Scale K k α
Results in Table 2.24 illustrate that both English and Urdu versions of FQS yield
companionship, help, security, closeness and conflict. Items on conflict subscale were
subscale) is also reversed scored. First of all data were reduced to factors using Principal
Component Analysis (PCA) with varimax rotation. For this purpose, sampling
Table 2.25
Kaiser-Meyer-Olkin Test of Sampling Adequacy and Bartlett Test of Sphericity for FQS
KMO was found to be .79 that exceeds the minimum value of .50 suggested by
(Field, 2005) and Bartlett’s Test of Sphericity yielded highly significant findings (χ2
(253) = 886.88, p < 0.001), which showed the suitability of the data for factor analysis.
152
Table 2.26
Principal Component Analysis with Varimax Rotation for FQS Urdu Version (N = 151).
Table 2.26 depicts the factor loadings of 23 items from FQS Urdu version. These
items measure the positive and negative aspects of friendship quality. The analysis
showed high communalities for all the items. A principal component analysis with
varimax rotation resulted in 6 factor solution that converged in 7 iterations and loadings
ranged from .37 to .83. The first factor included items (20, 14, 23, 18, 19 & 9) indicating
the aspects of help and sharing and intimacy. Item 9 loaded on both first and fourth
factors. This factor had the highest Eigen value (5.9) and accounted for the largest
proportion (23%) of the total variance. Items representing the conflict in friendship (15,
13, 10 & 17) made the second which accounted for of 10.73% variance with an Eigen
value of 2.68. One of the items (Item No. 17) also loaded on 4th factor; however the
difference was large enough to consider it integrated in 2nd factor. Rest of the matrix was
rather ambiguous with many cross loadings and factors explaining less than 6% variance
for the measure. Eigen values of these factors merely crossed 1. This mixed picture was
Keeping in view the vague factor structure obtained by PCA and assumed
relationship of latent variables, true factor analyzing models (principal axis, maximum-
likelihood, unweighted least squares, etc.) were supposed to be used. In case of fairly
normally distributed data, maximum likelihood is the top preference. If the assumption of
multivariate normality is violated then principal factor methods (i.e. principal axis factors
in SPSS) are suggested (Fabrigar et al., 1999). Therefore, data were first checked for
normality assumption.
154
Table 2.27
Friendship
Quality 51-105 83.755 (.92) .981*
*p < .05.
Quality of friendship total score ranged from 51 to 105 with a mean score of
83.75 and standard deviation was equal to .92. Shapiro-Wilk test that is recommended to
normality, principal factor axis with promax method of rotation was used. It is a
combination method that first provides orthogonally rotated (varimax) solution followed
by oblique rotation, resulting in more precise solution. This Exploratory Factor Analysis
produces the basic factor matrix and the pattern matrix. Findings are presented below.
155
Table 2.28
Factor Matrix for FQS Urdu using Principal Axis Factoring (N = 151)
Factor Loadings
Table 2.29 shows that proximal factor analysis resulted in an obvious 2 factor
solution. These two factors represent the two basic features of friendship quality. Items
that originally represent companionship, help, closeness and security are combined as
positive indicators of friendship quality that can be named as friendship support. There
156
are only two cross loadings (item 3 & item 6), yet the first one barely met the loading
criteria (.35) and the second one yields a negative value on 4th factor that contains no
other values. The second factor retained items closer to the original factor of FQS named
as conflict. This factor comprises all the items that are scored reversely. Item 16 did not
load on any of the factors, whereas item 1 has shown cross loadings and appears to have
an insignificant contribution.
Now the results of pattern matrix are presented that are obtained after promax
rotation.
157
Table 2.29
Pattern Matrix of Principal Axis Factoring with Promax Rotation for FQS Urdu Version
The pattern matrix reveals 6 factors, of which 3 factors can be clearly extracted as
the basic model of friendship quality among Pakistani children and preadolescents. Factor
4 and 6 have only two items each, yet personify the important aspect of the original
model that is companionship. Hence the first factor is a blend of items from closeness,
security and help subscales of original FQS. It can be named as Support. The second
factor is also a mixture of 3 items from closeness (6, 7 & 8), two items from help (3 & 5)
and surprisingly one item from conflict subscale (21). Yet the loading for this item is not
so high and contains a negative value. These items jointly portray the aspect of
acknowledgment of each other’s’ presence, and help and hence named as Intimacy. The
third factor typically retained items of the Conflict dimension of friendship quality model
by Bukowski, et al. (1994). Items featuring companionship did not load on a single
factor, rather distributed across two factors (4 & 6). However, only factor 4 can be
retained for having higher factor loadings and more meaningful grouping. Item 22 also
theoretical model posited by original Friendship Qualities Scale. However the distinction
between positive and negative attributes was pretty clear which provides support for
of FQS Urdu version is provided by explaining the items, total scale and subscales’
Table 2.30
Table 2.30 illustrates that items depicting conflict dimension and scored in reverse
had lower mean scores as compared to other items. Besides these, item 1 and 16 also had
160
lower mean values. Item to total correlation was very low for items 16 (r = .05), 17 (r =
.01) and item number 21 (r = -.03). It suggests elimination of these items as it would also
Cronbach’s alpha coefficients were obtained for the set of items obtained from factor
matrix (21 items; 2 subscales) and the pattern matrix (20 items; 4 subscales).
Table 2.31
Reliability Analysis of the Friendship Qualities Scale (FQS) Urdu Version and Subscales
The results from rotated factor analysis demonstrate high to moderate internal
consistency for Support (α = .78) and the total score (α = .76). Despite having 2 items
reliability estimates were found for Conflict (α = .59) and Intimacy (α = .49) subscales.
161
Keeping in view the above mentioned results of exploratory factor analysis and
item statistics, we decided to exclude item numbers 4, 16 and 22 from the Final Urdu
version of FQS. Thus the final translated scale appeared to be a 20 item measure of
friendship quality having good internal consistency (α = .76). Smaller sample size could
be the reason for such distinct factor pattern. These factors may yield better reliability
Two subscales of children’s social competence scale were used. Those were
Empathy and Helping Behavior, and Self-control/aggression. Data was available for 135
Table 2.32
Version (n = 135)
1 2 3 4 5 6
Conflict - - - - - -.04
Sc-Agg - - - - - -
Table 2.32 shows that all the positive aspects of friendship quality, support,
intiumacy and companionship were significantly correlated with empathy and helping
behavior. They were also significantly interrelated. However, all of these subscales did
not demonstrate relationship with conflict subscale despite being scored in reverse and
portraying the component of conflict resolution. The results afford support to the
construct validity of the measure. Moreover support for convergent validity was found as
behavior subscale. Rather an inverse relationship was found demonstrating that positive
features of friendship quality were associated with lower levels of aggression and more
self-control.
163
Table 2.33
Table 2.33 shows that none of the values were missing however extremely low
values were observed for most of the data. Further analysis was done to check normal
distribution of data.
164
Table 2.34
Skewness Kurtosis
N Min Max M SD Statistic SE Statistic SE
f1 672 1.00 5.00 4.0119 1.29880 -1.033 .094 -.241 .188
f2 672 1.00 5.00 4.1994 1.15896 -1.408 .094 1.015 .188
f3 672 1.00 5.00 4.3274 1.13866 -1.936 .094 3.286 .188
f5 672 1.00 5.00 4.3438 1.03558 -1.734 .094 2.477 .188
f6 672 1.00 5.00 4.1845 1.18304 -1.347 .094 .762 .188
f7 672 1.00 5.00 4.4107 1.03489 -1.799 .094 2.347 .188
f8 672 1.00 5.00 4.1116 1.13908 -1.221 .094 .778 .188
f9 672 1.00 5.00 4.3676 1.01422 -1.773 .094 2.683 .188
f10 672 1.00 5.00 3.3080 1.57440 -.326 .094 -1.442 .188
f11 672 1.00 5.00 4.4048 1.02713 -1.898 .094 3.120 .188
f12 672 1.00 5.00 3.9583 1.34332 -1.026 .094 -.305 .188
f13 672 1.00 5.00 3.6057 1.59568 -.617 .094 -1.256 .188
f14 672 1.00 5.00 4.2381 1.22316 -1.546 .094 1.270 .188
f15 672 1.00 5.00 2.5253 1.65666 -.449 .094 -1.479 .188
f17 672 1.00 5.00 3.1443 1.63831 -.131 .094 -1.598 .188
f18 672 1.00 5.00 4.2098 1.20568 -1.488 .094 1.130 .188
f19 672 1.00 5.00 4.1042 1.19407 -1.234 .094 .554 .188
f20 672 1.00 5.00 4.4315 .96962 -1.857 .094 3.096 .188
f21 672 1.00 5.00 2.9673 1.56493 .075 .094 -1.505 .188
f22 672 1.00 5.00 4.4375 1.08188 -2.055 .094 3.452 .188
f23 672 1.00 5.00 4.6131 .92281 -2.708 .094 7.172 .188
Table 2.34 illustrates that all the data on all the items of Friendship Qualities
Scale was not problematic in terms of normal distribution as the skew and kurtosis values
did not exceed absolute values of 3.00 and 8.00 respectively (Kline, 1998; 2011).
165
The CFA. We tested the 4 factor model which yielded acceptable values for
model fitting, yet the issue of multicollinearity was observed there for two factors. It is
important to note that reversed scored item 21 (from original conflict scale) which
actually loaded on factor 2 (Intimacy) with negative value in EFA pattern matrix, was
Although acceptable model fit was obtained (CFI = .92; TLI = .90), yet keeping in
view the existing evidence and suggestions (Kline, 2005; Schumckar & Lomax, 2004),
we re-specified the model and collapsed the factor 2 and 4 (See App. C). In view of
modification indices, three items (11, 14 and 3) which previously embedded in support
and intimacy were separated to create a new factor. It reduced the multicollinearity
among the factors. Additionally, errors of item 11 and 14 were allowed to covary as
suggested by Lagranian multiplier (Bentler & Dijkstra, 1985; Garson, 2015). Items 2 and
12 were also eliminated because of nuisance it used to create in the model (suggested by
modification indices). Item After these adjustments we ran CFA (Figure 2.4). It
improved the result and model fitted well to the data (CFI = .95; TLI = .94). Hence we
renamed the factors as items for support were separated as another factor, the rest of
items were named as approval. Finally approval, intimacy, help and conflict constituted
the Friendship Qualities Scale. The details of fit indices and the final model are presented
below.
166
Figure 2.4
Table 2.35
Table 2.35 clearly illustrates that model 2 was an excellent fit to the data as
compared with model 1 (also see App. C for model 1 and modification indices). Allowing
item numbers 14 and 11 to correlate, increased the CFI and TLI values, decreased
CMIN/DF and RMSEA. Additionally RMR was found .06 for both models. Goodness of
fit index (GFI) also increased from .94 to .96 (model 1 and 2 respectively).
168
Table 2.36
Item loaded high on than their respective factors. FQS total score was computed
by adding the scores of approval, intimacy and help. High score on combination of these
factors was regarded as good quality of friendship. Conflict scale being distinctively apart
was treated separately. Internal consistency and inter-correlations among factors and the
FQS total scores was computed. Results are presented in the following tables.
169
Table 2.37
K M SD α
Internal consistency of the subscales was within acceptable range as all the alpha
values exceeded .60 (Nunnally, 1967). Mean score for conflict scale was lower than
others. Reliability of FQS total score measuring positive aspects of friendship quality was
high (α = .80).
170
Table 2.38
All the three factors were highly correlated with the total score on FQS. Conflict
score was not significantly correlated to any of the other factors or the total score. Among
other three factors, highly significant correlations were found. Yet all of these were
below .60, which indicates they were uniquely measuring separate dimensions of
friendship quality.
171
Discussions
Urdu version was obtained using standardized translation procedure and the no in-
depth investigation was required as with the case of OBVQ. Results from bilingual
versions of FQS. We shall discuss the results of EFA, CFA and other psychometric
Exploratory Factor Analysis. The study explored the factor structure of FQS
after translating it into Urdu. The measure has already been translated and adapted into
different languages; it was considered important for exploring significant aspect of peer
bilingual sample yielded strong relationship between items of both versions which
affirms the language equivalence. Internal consistency of both the Urdu and English
Seven items showed cross loadings and the internal structure failed to convey some
meaningful outcome. For that reason, exploratory factor analysis was conducted using
principal axial factoring as PCA does not account for underlying correlation of latent
variables. Analysis was also performed with promax rotation and results of both factor
and pattern matrices were compared. The structure indicated the presence of a two main
factors. Items loading on these factors depicted positive and negative dimension of
friendship quality. Twenty three percent of the variance was accounted for by positive
friendship quality named as Support. The findings are in line with the Italian version of
172
FQS which also found Affect and Conflict dimensions as main factors pertaining to
nature of friendship quality. Three factors explained the positive features of friendship
quality and were named as Support, Intimacy and Companionship. Conflict exhibited the
same structure as illustrated in the original version (Bukowski et al., 1994). This pattern
is slightly different than those reported in most of the studies that supported the original
five factor model (Atik et al., 2014). Findings are yet consistent with some of the
Bukowski ‘s (1991) model and yielded similar kind of findings. Friendship quality was
The coefficients for factors resulted from pattern matrix were mostly low except
for Support (.78). These values are comparable to those obtained in the original scale
development study (Bukowski et al., 1994) and the Italian version (Fonzi, et al., 1997).
Reliability of the scale increased after deleting item 16 which did not distinctively load
on a factor and also showed poor item-total correlation. The Turkish adaptation study
also excluded this item (Atik, et al., 2014). The item is about going each other’s’ house
during free time. The reason for being inappropriate is evident considering our cultural
scenario. Children especially girls of this age are usually not allowed by parents to visit
each other’s’ house. Friend may not necessarily live in neighborhoods and hence it is
difficult to meet after school hours. Items 15, 17 and 21 also showed poor item-total
173
correlation (< .20). However considering the importance of the items and appropriate
factor loadings, these items were retained for final Urdu version of FQS.
divergent validity of the measure. Support, Intimacy and Companionship resulting from
rotated matrix were significantly correlated with empathy and helping behavior while
unrelated to empathy and helping behavior and also with the positive friendship features
(Support). Results offer adequate evidence for the construct validity of the FQS Urdu
finally determine the soundness of psychometric properties of the Urdu version. Earlier
findings have also established connections between friendship quality and empathy as
both are important elements of interpersonal competence among youth (Chow, Ruhl, &
Buhrmester, 2013). French, Lee and Pidada (2006) have discussed the disparities in the
perception friendship quality construct and its assessment and in their work on peer
and unreliable source of information in studies of social development (Patterson, Reid &
Dishion, 1992). Using multiple methods and multiple sources may resolve the issue.
It is difficult to draw to final conclusions about the factor structure of the FQS
Urdu version. Both of the proposed models are subjected to further evaluation with lager
sample and confirmatory factor analysis. We retained 4 factor model for further
evaluation It affords more parsimony (Sethi & King, 1991), though consistency and
interpretability (Leader & Sethi, 1992) are compromised somewhat. Finally, using a 20
174
item Urdu version of Friendship Qualities Scale was considered suitable to use for the
main study.
In the final stage of this section, we executed CFA to confirm the EFA four factor
model for the data of main study. Results of confirmatory factor analysis initially
provided support to exploratively obtained model of friendship quality. Though CFI and
TLI values fell within acceptable ranges of a less stringent criterion (>.90) however
modification indices suggested valuable changes which could decrease the CMIN/DF and
increase the baseline comparison indices (Schreiber et al.., 2006). Employing these
suggestions we added a new variable (support) for strong theoretical reasons. Three items
(3, 11 & 14) belonged to the subscale Help from original FQS model, and covariance was
also found amongst them. Doing so increased the construct validity of the measurement
model and decreased the multicollinearity found among factors in model 1. Lagranian
multiplier further suggested some covariance in error parameters of item numbers 2 and
12 with items in conflict scale. We incorporated the suggestion and found notable
changes in fit values. Re-specified model provided an excellent fit (CFI = .95; TLI = .94,
RMSEA = .03, CMIN/DF = 1.7). All the factor loadings were greater than .35 and
reasonable covariance among factors was found (< .85; Kline, 2005) which indicated that
each factor distinctly measured different aspect of friendship quality. Although such
changes might compromise the parsimony of the model, however a strong theoretical
ground, cultural diversity of adapted version and significant increase in fit indices are
enough to explain the logic behind these adjustments (Schumcka & Lomax, 2004). We
carefully executed this process by not allowing correlations of errors from two different
factors and only larger indices were considered for change. Experts support such
175
revisions until a good fit is achieved. Moreover, there is evidence to combine the
categories if they demonstrate poor discriminant validity (as found in model 1) (Toe, Tsai
support and conflict. Factor other than conflict were strongly correlated and were found
either unrelated or inversely related with conflict scale. Therefore total score of
friendship quality was computed for three factors while conflict scale was treated
separately. Reliability estimates for all the subscales were acceptable (.60 to .70)
(Nunnally, 1967). Moreover it is not unusual to obtain poorer reliability and different
pattern of reverse scored items than normally scored items. Sometimes, respondents
overlook the negative item and maintain same response style. Excluding them or dealing
them separately is a better idea to maintain the reliability of the measure (Conard et al.,
2004). These variations in present study could be subjected to younger age group of
children and also the grade level. Students in 4th and 5th grade might not be mature
enough to resolve their conflicts, yet still appreciate the positive aspects of the
relationship. While students in 6th grade go through school transition phase and
friendships are still at early stages and later could become stronger.
The findings of study 1 present good support for the internal consistency and
Chapter-III
Section A. The first second section provides prevalence estimates of bullying and
peer victimization across gender and grades. Prevalence and differences for different
types of victimization and bullying across gender and grade were reported. Data on
remaining parts of OBVQ (i.e. severity, location, duration, place and reporting of
bullying, and students’ reactions toward bullying) were also evaluated to portray a
measures of academic performance that are exam marks and school attendance
percentages of the students. Differences are also observed for demographic variables such
sector schools.
students.
177
Hypotheses
grades.
4. Boys would report more direct form of bullying/victimization (e.g. verbal and
Method
Participants
For sample details, see sample III, phase 2, section A of the study 1. Demographic
Measures
psychometric properties with being bullied and bullying others dimensions yielding
Chronbach Alpha of 0.79 and 0.82 respectively. The concurrent validity using original
and the translated version with a sample of 36 bilingual students had also demonstrated
significant correlation coefficients ranging from 0.77 to 0.94. The current study further
evaluated the reliability and validity of the two sets of items measuring bullying and
Most of the earlier studies have used the cutoff point of ‘‘2 or 3 times a month’’
subsequent forms (Glew et al, 2005; Kristensen & Smith, 2003) as per recommendations
by Solberg and Olweus (2003). However some researchers have used a bit less strict
(Vieno, Gini & Santinello, 2011). This standard can be followed especially for cyber
bullying (Monks, Ortega-Ruiz & Rodr´ıguez-Hidalgo, 2008; Raskauskas & Stoltz, 2007)
However, we chose the cutoff point of ‘‘2 or 3 times a month’’ for all items except for
179
cyber bullying that followed the “once or twice a month”. So, for overall involvement
and each type of bullying a student was classified as (a) exclusively involved in bullying
others (bully only), (b) exclusively involved in being bullied (victim only), (c) involved
in both bullying others and being bullied (bully-victims), and (d) completely not involved
number of total family members and monthly household income was enquired (see App.
D).
Academic Record
Academic record included average of last two exams and average class attendance
of last two months. Most of the schools did not cooperate to provide this information. As
three schools totally refused to share the record while several reminders were sent to
other schools for providing the required details. Three schools (one private, one semi-
government and a public sector school) gave direct access to their record.
Procedure
Since the data were obtained from the same sample used previously, consult the
last segment (CFA Sample III) in 2nd phase of section A of the first study.
180
Results
Results for study 2 are presented by two sections. First section includes the
outcomes.
This section mainly aimed at describing prevalence of bullying and victimization, the
types, and other relevant information among bully/victim groups. Olweus Bully Victim
Questionnaire provides abundance of such information but more pertinent aspects have been
reported in the dissertation. For those who reported being bullied as victims or bully-victims,
data on class and number of bullies, severity and duration of bullying, the place where
bullying occurred and to whom they reported the occurrences of bullying was analyzed as
being valuable for intervention and prevention measures. On part of those who reported
bullying others as bullies and bully-victims, Data were analyzed for frequencies, percentages
Figure 3.1
The Figure (3.1) indicated that relatively smaller percentage of the total sample
acknowledged that they were just bullies, never victims (17.3%). Students classified as
victims were 19.3%. Most of the students were involved both in perpetration and
groups made 34.6% of the total sample indicating that 65% students were involved in
bullying behaviors.
182
Figure 3.2
Figure 3.2 showed that greater proportion of bullies (19.2%) was in 5th grade
while 14.6% belonged to 4th grade and 17.5% bullies were studying in 6th grade. Lowest
proportion of victims (16.9%) was also found in 5th grade. Students in grade 4 were most
often recognized as victims (23.9%), whereas 18.3% victims studied in grade 6. A large
proportion of bully-victim group belonged to 4th grade; while 26.8% studied in grade 5
and 2.1% bully-victims were students of grade 6. Uninvolved group constituted 25.4% 4th
graders, 37.1% 5th graders and 38.1% students studying in 6th grade.
183
Figure 3.3
Figure 3.3 illustrated that equal number of boys (17.2%) and girls (17.3%) were
identified as bullies. Slightly higher numbers of boys (20.2%) were classified as victims
as compared to girls (18.6%). More boys (31.2%) were classified as bully-victims than
girls (25.9%). Uninvolved group mostly included girls (38.2%) as compared to boys
(30.5%).
184
Table 3.1
Total 141 (17.3% 158 (19.3%) 235 (28.8%) 283 (34.6%) 817
Table 3.1 indicated that proportion of bullies was almost equal in public and
private schools (17.2% and 17.9% respectively). The ratio was relatively lower in semi-
government schools (16.1%). However victims were most often found in semi-
government schools (21.7%). Hence again public and private schools had minor
followed by private (25.9%) and semi-government schools (21%). Most of the students
whereas fewer students in public sector schools were identified as uninvolved (30.8%).
185
Table 3.2
Pearson Chi-Square for Bully-Victim Groups across Gender, Grades and type of Schools
χ2 df p
N of Cases 817
Chi-square test revealed that the relationship between gender and bully–victim
status type differences was non-significant. It showed that boys and girls were equally
distributed into bully-victim groups, χ2 (3) = 6.4, p = .09. Pearson chi square test for
groups by grade levels, χ2 (6) = 16.38, p < .01. Chi square results were non-significant
Following tables compare bullies and victims with bully-victim group for frequent
involvement in bullying behaviors. The standard cut off (two to three times a month) is
represented as occasional bullying. The last two response options (once a week, several
times a week) were categorized together as frequent bullying. This category represents
Table 3.3
f (%) f (%) N
Table 3.4
f (%) f (%) N
or more frequent bullying and victimization among perpetrators and victims in three
different roles did not differ significantly. Severity of bullying and victimization yielded
Follow analysis were only performed with regard to students who reported being
bullied either as victims or bully-victims. With reference to these 393 students (158 victims
and 235 bully-victims) who had certainly experienced some form of bullying in general,
further investigation was conducted to determine the frequency in which various incidents of
bullying occurred. Thus, the focus was the frequency of these incidents differed for the three
Table 3.5
Gender Grade
Called mean names, teased 124(32.8) 113(25.6) 69(33.7) 63(29.6) 105(26.3) 237(29)
The percentages for being bullied by types demonstrate the highest prevalence
estimate (29%) for verbal victimization that is characterized by name calling and teasing
in hurtful manner followed by being subject of rumor (24.2%) and excluded or ignored
by others (23%). Having money or possessions taken or damaged (14.2%) was least
reported. Thus direct verbal and relational types of peer victimization have been found
Table 3.6
Pearson Chi-square for the Types of Victimization across Gender and Grades
χ2 p χ2 p
Table 3.6 indicated that more boys than girls were teased and called mean names
with harm, χ2 (1) = 5.124, p < .05; were exposed to ethnic and racial comments, χ2 (1) =
4.37, p < .05, and cyber victimization, χ2 (1) = 24.93, p < .001. Students especially girls
190
were less frequently bullied by cyber means. Social exclusion that is typically
experienced by girls showed the same pattern for the current sample. Other types were
Verbal victimization seemed to be associated more with lower grade levels yet
this association was not found significant, χ2 (2) = .59, p < .166; while students in higher
grades tended to experience more physical victimization, χ2 (2) = 8.19, p < .01. Both of
the indicators of relational victimization i.e. exclusion and rumor were different across
grade levels, χ2 (2) = 22.67, p < .001and χ2 (2) = 22.78, p < .001. Furthermore damage to
property, χ2 (1) = 14.59, p < .001, racial χ2 (1) = 13.50, p < .01 and sexual victimization,
χ2 (1) = 18.92, p < .001 experiences also differed across grade levels. Students especially
girls were less frequently bullied by cyber means yet this variation was not significant, χ2
(1) = 1.75, p = ns
Moreover victims and bully-victim groups were compared for their experience of
various types of victimization. Chi square tests of independence were computed for this
purpose. The only significant difference between pure victim and bully-victim group was
found for being threatened, χ2 (1) = 11.75, p < .001 as almost 72% of victims within this
Who is bullying?
Item numbers14 and 16 were evaluated to check the information provided by victims’
about those who bully them. This portion is also includes only the responses of students
who mentioned that they had been bullied (victims and bully-victims).
191
Table 3.7
Gender Grade
Table 3.7 indicated that most of the victims were bullied by their own class fellows.
Boys and girls seem to differ in this regard. Inter class discords are more prominent among
boys while girls were mostly bullied by their own class mates. Grade differences can also be
observed. Within class victimization was reported mostly by students of 6th grade. Students
of 4th and 6th grades were more likely to be bullied by seniors and many of the 6th grade
Table 3.8
Number of Bullies Reported by Victims and Bully-Victims across Gender and Grades
Gender Grade
No. of Boy Girl 4th 5th 6th Total
Bullies
Only One 103(36.8%) 177(63.2%) 82(29.3%) 73(26.1%) 125(44.6%) 280
Table 3.8 demonstrated that most of the bullying was done by either a single
individual or the small group of 2 to 3 students. Single person bullying was more often
reported by girls while most of the boys experienced it by groups. Grade differences were
also noticeable as students in 6th grade outnumbered others in reporting about number of
Table 3.9
Pearson Chi-square of Bully’s Class and Number across Gender and Grade
χ2 p χ2 p
reporting about bully’s class and number were significant except for number of bullies
Duration of bullying
Table 3.10
Duration of Victimization Reported by the Total Sample and Those who had been Bullied
Groups 1-2 weeks About 1 month About 6 months 1 to several year Total
Bullying mostly lasted for one to two weeks (41.5 %) as reported by the total
Figure 3.4
60
50
40
30
20
10
Figure 3.4 showed that playgrounds (53.2%) and class room (45%) in the absence
of class teacher were the most sensitive areas in schools as most of the bullying was
reported to occur at these places. Bullying also occurred frequently at canteen and in the
195
class even in the presence of teacher. It rarely happened in the vehicle (bus, van, rikshaw)
Table 3.11
χ2 p χ2 p
As far as grade was concerned, significant differences were found regarding two
places where bullying takes place, namely playground and in class (teacher absent). Students
in 4th (61.8%) and 5th (53.8%) were more often bullied in playgrounds compared to grade 6
students (49.6%). Students in grade 6 were more frequently bullied in class in the absence of
playground only (60.4% boys; 45.9% girls), while no significant differences were found with
Disclosure of bullying
Table 3.12
Frequency Distribution of the Boys and Girls and Students of Different Grades about
Disclosure of Bullying
Grade Gender
Complain th th th Boys Girls
4 5 6 Total
f (%) f (%) f (%) f (%) f (%) f (%)
Did not Report 35(28.5) 19(20.4) 55(31) 63(32) 46 (23.5) 109(27.7)
Most of the students disclosed information about being bullied to someone (72.3%).
Still nearly one fourth (27.7%) did not share their experience with anyone. It was further
investigated whether, in terms of the gender and grade, the students who had been bullied
told someone about the incident(s). With regard to grade (χ2 (2) = 3.49, p = ns) and gender
Figure 3.5
50
45
40
35
30
25
20
15
10
0
Class teacher Parents school staff Sibling Friends
Figure 3.5 indicated that most of the students (46.1%) reported to the school staff
about their experience of victimization. Many of the students shared their experience of being
bullied with friends (32.8%) and also told the class teacher (32.3%). Siblings were the second
last preference of students about reporting that bullying occurred (24.7%). Fewer victims or
Data on OBVQ item 24 to 32a were analyzed to identify the most frequently
Table 3.13
Gender Grade
Called mean names, teased 133(35.3) 115(26.1) 62(30.2) 68(31.9) 118(29.6) 248(30.4)
type remained highest among all types of bullying followed by exclusion and physical
types. Boys were more verbally abusive compared to girls. Girls were typically more
199
involved in relational bullying i.e. excluding others and spreading rumors, as compared to
Pure bullies and bully-victim group were alike for involvement in most types of
bullying except for the occurrences of threatening and forced action that were
significantly higher, χ2 (1) = 14.15, p < .001, in bully-victim group (79.1%) as compared
about Bullying
This portion analyzed responses on different items of OBVQ with the help of
independent sample t-test to check the gender differences and one way ANOVA to
Table 3.14
95% CI
Items t df LL UL Cohen’s d
1. School liking -2.845** 815 -.332 -.061 -.198
The Table 3.14 demonstrated significant gender differences for school liking, t
(815) = -2.845, p < .01; empathy towards victims, t (815) = -2.036, p < .05; willingness to
join bullying, t (815) = -2.108, p < .05; and students’ own reaction towards the act of
bullying, t (815) = -2.310, p < .05. Girls (M = 4.5, SD = .89) tend to like their school
more than boys (M = 4.3; SD = 1.07). Both boys and girls had almost equal number of
empathy, joining in bullying and own reaction indicates pro-victim attitude. Girls showed
more pro-victim attitude as compared to boys. Boys and girls did not differ in being
afraid of bullying at school and their perception of class teacher’s efforts to stop bullying.
201
Table 3.15
School liking 4.45 (1.04) 4.44 (1.03) 4.44 (.94) .019 2 .009 .010 .990
No. of Friends 3.45 (1.19) 3.42 (1.16) 3.27 (1.12) 4.531 2 2.266 1.581 .206
Empathy 3.51 (.91) 3.43 (.97) 3.39 (1.01) 2.154 2 1.077 1.123 .326
Join bullying 3.83 (1.71) 3.91 (1.83) 3.81 (1.85) .999 2 .499 .152 .859
Own reaction 3.81 (2.24) 3.90 (2.06) 4.13 (2.14) 16.352 2 8.176 1.768 .171
Being Afraid 2.30 (1.5) 2.02 (1.37) 2.20 (2.09) 8.880 2 4.440 1.379 .252
Class teacher 3.22 (1.62) 3.23 (1.63) 3.39 (1.57) 4.002 2 2.001 .778 .460
Table 3.16
One-Way ANOVA Showing Mean Differences on OBVQ Items (N =817) across Bully-
Victim Groups
School liking 4.41(.97) 4.41(1.1) 4.30(1.1) 4.60(.83) 12.01 3 4.01 4.13 .006
No. of Friends 3.48(1.2) 3.12(1.3) 3.24(1.2) 3.46(1.1) 16.37 3 5.46 3.84 .010
Joining bullying 3.29(1.7) 3.86(1.8) 3.38(1.6) 4.48(1.7) 207.59 3 69.19 22.73 .001
Own reaction 4.19(2) 3.55(2.3) 4.05(2.1) 4.10(2.1) 40.69 3 13.56 2.95 .032
Being Afraid 2.17(1.5) 2.19(1.4) 2.89(1.6) 1.66(2.1) 166.68 3 55.56 18.34 .001
Class teacher 3.44(1.6) 3.06(1.7) 3.17(1.5) 3.56(1.6) 35.05 3 11.68 4.60 .003
.83), greater number of friends (M = 4.60, SD = .83), more liking for school (M = 4.60,
SD = .83) and was less afraid of bullying at school (M = 4.60, SD = .83). They also
= .83).
203
Table 3.17
Means, Standard Deviation, Skewness and Kurtosis for Academic and Demographic
All variables were approximately normally distributed except for monthly income
which presented great variation ranging from 4000 to 400000. Variation was also found
in family size.
All these variables were not coded and were treated as continuous variables.
Being outcome of bullying, academic variables (marks percentage and attendance) were
evaluated for differences across gender and bully/victim groups using 4X2 Univariate
siblings and birth order were analyzed for association with victimization (OBVQ item 4;
Table 3.18
Dependent
Source Variable SS df MS F p η2
There was no significant main effect of gender on exam marks percentage and
was demonstrated F (3, 355) = 14.83, p < .001, η2 = .11. Gender and group interaction
Figure 3.6
Table 3.19
Post Hoc Scheffe Test for Multiple Comparisons across Bully/victim Groups for
Academic Achievement
95% CI
(I) Bully/Victim (J) Bully/Victim
All the bully/victim groups significantly differed than uninvolved group in their
bully-victim group showing the least scores, though the difference was found
Though there were no significant differences among bully/victim groups for their
average monthly school attendance (see table 3.19), however the profile presents quite
insightful picture.
Figure 3.7
Figure 3.7 depicts that victims followed by bully-victim group had relatively
lower attendance percentages as compared to bullies and uninvolved student whose ratios
were alike.
Table 3.20
Table 3.21
Father’s Mother’s
Education Education
Educational Level f % f %
Data was coded and parents who never went to school were categorized as
illiterate, those having education up to or below 8th grade were labeled as middle and
below. Those who completed more than middle standard yet less than 14 years of
education were put under the category of secondary and higher secondary. Those who
completed more than 14 years of education were classified into group of graduation and
above. Father’s level of education was better than mothers’ educational level. Data were
fairly normally distributed across sample. Only 41.1% participants provided this
information.
210
and victimization was explored using MANOVA. Two global items representing
Table 3.22
Dependent
Source Variable SS df MS F p η2
Discussion
The findings for prevalence study showed that 65% students were involved in
some kind or another in bullying. Amongst them 17.3% reported being a bully, 19.3%
reported being the victim, while a larger proportion (28.8%) admitted that they were
involved in bullying others and were also victimized by their peers, constituting the
compared with prevalence reported in other studies. Nansel, et al. (2001) found that as
much as thirty percent of students had been influenced by bullying in some way.
Panayiotis et al. (2010) used Cyprotic version of OBVQ and reported a rate of 17% as
involved in bullying. Some recent reports across the world found 5 to 45 % students
involved in different roles (Laeheem & Baka, 2009; Kashirsagar et al., 2007). Yet these
results are comparable to large scale studies (UNICEF, 2012) according to which low
income regions (e.g. African countries) were at greater risk (20% to more than 80%) of
is suggested as the whole class approach was not followed. Additionally, prevalence rates
current sample suggests that most of the students were at high risk for harmful outcomes
found for gender or types of schools. Grade differences emerged as students in lower
grades (4th) were mostly categorized as bully-victims and victims than other two
justification for result of grade 4. Being senior most in elementary section, students of
grade 5 are less likely to be bullied, while more likely to become perpetrators and their
targets could be students of grade 4, who either become submissive victims or reactive
ones (Frey et al., 2005). . Another reason could be more truthful reporting by younger
students who are less subjected to demand characteristics. This trend of lower reported
victimization in higher grades is consistent with earlier reports (Selekman & Vessey,
2004).
last two response alternatives of two global items corresponding bullying and
victimization. Comparisons across bullying roles, gender and grade did not find any
students. Yet it was found that students reported more exposure to frequent victimization
than bullying. Experience of being bullied usually lasted for one or two weeks followed
by the duration of one month. More than forty percent students have been experiencing
victimization for more than 6 months that adds to the sensitivity and seriousness of the
efforts.
213
Prevalence for the types of bullying revealed findings consistent with the previous
ones as verbal bullying and victimization were the most commonly reported form across
genders and also grades (especially elementary grades) (Coloroso, 2004). For the present
sample, verbal bullying accounted for 29% of the reported victimization. Rumor
spreading and excluding from group were also found common as more girls and students
in grade 6 reported being bullied by these means. The results support the earlier studies
(Baldry, 2004). Boys reported being bullied by cyber means more often than girls. It was
also found that bully-victim group reported being threatened more as compared with
victims. These threats might evoke them to react aggressively. With reference to bullying
others, again high prevalence was found for verbal bullying followed by social exclusion
and physical bullying, having boys involved more in both of the direct modes (verbal and
physical) and girls typically were found involved in relational types (excluding others and
rumor spreading) (Underwood & Rosen, 2011). In accordance with the finding about
victimization, bully-victim group was involved more in threatening others and also
Most of the bullying took place in small group of 2 to 3 persons was done by a
single individual. Bullying by group was more common among boys and students in
higher grades were bullied by larger groups. Bullying was most likely to occur in less
supervised area such as play ground (53 %) as reported in earlier studies (Craig, Pepler,
& Atlas, 2000), however occurrence rate in classroom was also high (45%). Students
mostly reported the incidents to school staff friends and parents. High prevalence
indicates that the issue has not been taken seriously and friends were less able to lend
214
support to the victimized peer. Additionally, girls showed more pro-victim attitude
compared with boys (Almeida, Caurcel, Machado, 2006; Rigby & Slee, 1991).
Findings mostly adhere to the existing evidence regarding grade and gender
differences with some exception (Carlyle & Steinman, 2007). An overall decrease has
been observed in rates of bullying especially by physical means (Molcho et al, 2009).
Timely efforts and awareness can produce fruitful results in decreasing bullying.
Most of the data on academic and socio-economic variables was found missing.
From the available data set we evaluated average percentage of last two (consecutive)
exams and average percentage of class attendance (last two months) as academic
bullying adversely affected percentage of marks in exams. Bullies, victims and bully-
victims were equally affected and performed significantly poor in exams compared with
school attendance, yet the tendency of being absent from school was more prominent
among victim and bully-victim groups. Academic cost of bullying has been reported in
None of the socioeconomic variables such as monthly income, family size, birth
order, parent’s education was found associated with bullying. Findings are conflicting to
(Shujja et al., 2014). Findings also oppose the stereotype that bullying is associated with
socioeconomic status and victims come from economically disadvantaged families or less
educated parents (Bowes et al., 2009; Jansen et al., 2012; Nordhagen, Nielsen, Stigum,
215
bullying and socioeconomic status (Tippett & Wolke, 2014). However the current result
could be subjected to the assessment of these variables. As it was not the main focus of
the study, we did not determine any socioeconomic classes; rather data were treated as
continuous variable and some important links could have been missing. Moreover
monthly income reported for the sample varied a lot. It restricts the generalizability of the
findings.
216
Chapter-IV
explaining bullying and victimization. Several factors and aspects have also been
discussed in detail. Researchers have agreed that bullying behavior could be influenced
have been acknowledged as the essential elements accounting for the process of bullying
(Cook et al., 2010; Woods et al., 2009). In this view the present research empirically
recognition and the quality of friendship have been examined in two studies (study 3 and
4 respectively). This chapter provides an account of both the studies and integrates the
findings in the end with regard to demographic trends and academic cost of poor
Section A. Study 3: Effect of mood on Emotion Recognition among School bullies and
victims (Study 3)
Victims.
Section A: Study 3
One of the foremost objectives of the main study was to determine the effect of
standardized mood induction procedure was required appropriate for the participants’
age. Extensive literature review provided support to several mood induction procedures
being widely used with children and early adolescents; autobiographical recall, story or
vignette based mood induction following Velten procedure, music mood induction, and
mood induction via video are to name a few (Brenner, 2000). Web-based pictures with
emotional content have also emerged to be effective in inducting mood states (Goritz,
2007; Goritz & Moser, 2006). Keeping in view the time constrains and availability of
materials in accordance with Pakistani culture, we opted for a tryout of four methods of
mood induction (i.e. Instrumental music, Video songs, Emotive pictures and a
phase examined the magnitude of change in valence and arousal reported after mood
Current study was a prerequisite for the main study. It was divided into two parts.
Firstly, mood suggestive materials pertaining to all the four mood induction techniques
(instrumental music, video song, emotive pictures and pictures with music combination)
were collected and rated by the judges for the magnitude of positive or negative affect
they could produce. Materials rated high on emotionality in respective mood dimensions
218
were selected for final evaluation by a sample of 64 children. In the next step
effectiveness of these techniques was evaluated in inducing the desired mood (i.e. happy,
Hypotheses
Video song would be more effective in suggesting the desired mood as compared to three
other techniques.
Method
Participants
Total 64 participants (Mage = 10.79; SD = 1.08; 55 boys) from one private and two
public sector schools (one girls school, one boys school) participated in this study.
Initially 72 students were randomly selected from the attendance list of grades 4th (n =
24), 5th (n = 24) and 6th (n = 24). Four of them refused to participate. Three participants
were not available at the given time due to participation in sports, and one student
discontinued during the experiment. Thus the final sample of the present study consisted
of 64 participants from 4th (n = 20), 5th (n = 23) and 6th (n = 21) grades.
Measures
was used for assessing valence and arousal at baseline and after each induction. It
of mood (i.e. pleasure, arousal and dominance). It is a 9 point Likert type scale
presenting five pictures with four spaces in between. The pictures are cartoon type
changing to an unhappy figure, and arousal dimension ranges from an excited, wide-eyed
figure to a relaxed, sleepy figure. In the current study, paper and pencil version of two
scales were used. SAM figures used in present study progressed in valence and arousal
from right to left direction. Hence the highest valence/arousal is scored as 1and the
lowest as 9. Thus the scale rated for valence scoring from one (happy) to nine (sad) and
the second scale, rated for arousal being scored from one (calm) to nine (excited).
However this may not affect the results necessarily and similar scoring has also been used
earlier (Ong, Carde, Gross & Manber, 2011). The procedure was self-paced as there was
no time limit for rating. Earlier studies have successfully used the SAM with child
samples (Backs, Silva & Han, 2005; Sharp, Goosen & Goodyer, 2006). SAM has
typically been used in combination with International Affective Picture System (IAPS;
Lang, Bradley, & Cuthbert, 1997) for rating emotive pictures. It has also been widely
used for rating other mood and emotion evoking materials such as sounds or videos for
Hesse, 1994; Westermann, Spies, Stahl, & Hesse,. 1996) have confirmed the superiority
instrumental music and music with lyrics and visual effects had been effectively used in
different mood induction experiments (Fairclough, van der Zwaag, Spiridon, &
induction was available within our cultural context, different options were evaluated
The material for this technique was finalized in two steps. The first step included the
selection of audio, video and instrumental music. These selections were then rated by
sample of students from the target population of the main study using an 11 point single
item Ottawa Georgia Mood Scale (Cheng & Ward, 2005). It ranges from zero (extremely
Music selection. The first step involved the preselecting of a piece of music for use
with all subjects to ensure its effectiveness at eliciting sad and happy mood for all the
Carlsson, Hilmersson, & Juslin, 2009; Paquette, Peretz, & Belin, 2013). Three individuals
(all doing PhD in Psychology) participated in producing the list of music excerpts
including audio/video songs and instrumental music. Five video songs referring to a
happy mood state and five indicating sad mood were listed on the basis of perceived
effect. The list was developed keeping the age range of the participants in view and
considering all ethical issues as the videos were based on different movie songs. All
Five happy and five sad songs without video were also included in the trial. Lists
were produced in same manner. Similarly five pieces of music (instrumental) pertaining
to happy mood state and five instrumental music pieces eliciting sad mood were also
221
listed after extensive internet surfing. Each musical excerpt and audio songs were
downloaded from the Internet in the form of an MP3 file that was converted into a
standard audio format and recorded onto a CD for use in a standard CD player. All the
music was cheerful, familiar and age-appropriate as compared to the typical classical
music used with adults in the light of recommendation by Schellenberg, Nakata, Hunter,
and Tamoto (2007). Moreover a neutral instrumental tone (downloaded from internet,
Anonymus, n.d) that was mutually selected by members in selection committee was also
Music ratings. Thirty children (50% boys; Mage = 10.56, SD = 1.04) drawn from a
private evening tuition center were divided into two groups based on random allocation
(according to 2 types of music; video song, instrumental music). Each participant in both
groups (song and instrumental) was presented with sad and happy music excerpts in a
counter balanced order. Neutral excerpt was included in instrumental condition only. The
stimuli were presented one at a time through the headphones. They listened to these
preselected pieces of video, and instrumental music and rated them according to how
likely that piece of music could help people to feel sad or happy on an 11 point scale
ranging from 0 (extremely sad) to 10 (extremely happy). A neutral mood was represented
by the central point (5). Music was played at an appropriate volume, which subjects could
hear clearly and comfortably. The list of five music excerpts was presented to them by
asking them to listen to at least 1 minute of each piece. The pieces of music were,
however, long so they were allowed to move onto the next piece at any point after a
minute. Two video songs and two instrumental music excerpts, each representing happy
222
or sad moods and the neutral piece of instrumental music were finalized on the basis of
instructions for getting into the target mood. Studies have shown the effectiveness of
visual images in inducing desired mood (Brunyé, Mahoney, Augustyn, & Taylor, 2009;
Lang, Bradley, & Uthbert, 1995). Pictures suggesting happy, sad and neutral mood were
or depressing (sad) effect on the viewer, were gathered from the internet. Happy set of
pictures was based on the pleasant themes like happy children, attachment with parents,
play, fun and enjoyment, cuddly animals, leisure time, success at sports or exams,
celebrations and festivals. The sad set included unpleasant scenes such as destruction,
disasters and victims, crying children, poverty, accidents, grieved people, illness etc.
Neutral pictures illustrated non-emotive scenes for example silent crowd, traffic on the
road, people busy in routine affairs, buildings, house hold goods etc.
Picture ratings. In a pilot test, 10 judges rated these pictures on an 11 point Likert
scale with 0 indicating extremely sad, 5 as neutral and 10 as extremely happy. Two
separate orders of these pictures were constructed to balance order of presentation. Each
subject was randomly assigned to view pictures in one of these orders. For ethical
reasons, people assigned to rate the negative photographs were informed beforehand that
some of the photographs were graphic, and if they did not want to see the negative
photographs they may shift to positive pictures. Yet none of the judges did so.
223
Participants viewed the pictures as a power point presentation. No time limit was fixed
and they were allowed to see each picture as long as they wanted before rating. Finally 30
pictures with the lowest mean score were used for the sad and the 30 with the highest
score for happy mood induction. Pictures with mean scores around 5 were included in
neutral set. The neutral clip did not generate any emotions to a notable degree. For ethical
reasons, 3 sad pictures marked as extremely sad by at least 4 judges were excluded from
Previously, Baumgartner, Lutz, Schmidt, and Jäncke (2006) explored how musical
stimuli could improve the feeling of affective pictures. Mayer, Allen, and Beauregard
(1995) also employed this combination successfully in their mood induction study.
Following the same method, pictures were combined with congruent emotional musical
excerpts (instrumental pieces). The set of happy, sad and neutral pictures and
Experimental Design
The study followed a mixed factorial design with 7 trials repeated within 2
Procedure
After obtaining baseline mood and arousal ratings on SAM, participants were
randomly assigned to one of the two mood groups named as Positive Mood Induction
(PMI) group and Negative Mood Induction (NMI) group. During experimental session,
224
participants in each group received seven trials of inductions using four induction
techniques (i.e. instrumental music, video song, pictorial stimuli and pictures with
background music combination). At first they were primed to induce the desired mood
state (e.g., happy or sad), followed by neutral induction to reinstate their mood back to
neutral.
Total four randomized blocks (PMI-A, PMI-B, NMI-A & NMI-B) were created
for positive and negative MI group. In positive MI group, block A was presented with
instrumental music first (Trial 1 & 2) followed by video song (Trial 3), pictures (Trial 4
& 5) and the combination technique (Trial 6 & 7). Block B saw the emotive pictures
first, then video song, instrumental music and combination of pictures and music at the
The neutral induction was incorporated in each block for serving as a control
condition for each mood induction procedure and to restoring participants’ mood state
back to neutral before the next induction technique was administered to avoid ceiling
effects. Additionally, it provided a pre-induction rating of mood for the next induction
technique, and also ensured that participants might not leave the room in an altered mood
state. Identical instructions were given to happy, sad, and neutral inductions to account
for demand characteristics. It is important to note that ratings on an 11 point Likert scale
were obtained for affective valence and arousal of mood after each induction.
Figure 4.1a
(N = 64)
(n = 32) (n = 32)
The files for blocks were already saved on the desk top in folders. All the
participants were able to use the laptop either due to earlier familiarity and practice or the
current practice trial, therefore none of the participants reported any difficulty in handling
it. They were verbally instructed upon their arrival at the experimental room. Mood
rating was obtained after practice trial. Relevant folder for the block was open when they
sat in front of the screen. Following the instructions, they opened files one by one and
received inductions. Ratings were followed after each induction. Participants listened to
all kinds of music via headphones, whereas pictorial induction was administered via a
slide every seven seconds to ensure exposure to each picture for the allotted time. In
video song and combination techniques, they wore headphones and watched the visuals
on the screen.
experiment to deal with the sad mood carry over effects. One packet of potato chips and a
chocolate was given to each participant as a token of thanks at the end of experiment.
All four affect induction methods were successful in inducing three mood states
(i.e. happy, sad and neutral) pleasant and unpleasant affective states. The viewing image
with music and video song were most effective in enhancing the affect.
227
Results
Results of the present study are divided into two sections. First section depicts the
ratings obtained from judges and children for different mood priming materials i.e. video
songs, instrumental music and mood suggestive pictures. Descriptive statistics for these
ratings was computed to select the most suitable and effective pieces of music and set of
pictures for sad, happy and neutral mood conditions. Second section presents results of
sample. For analysis, the baseline ratings were entered as the first trial serving as pre-
induction for the first MI technique. Thus a 4 × 2 within subject factor resulted in total 8
combinations (1 baseline and the 7 induction trials). Two blocks (A & B) in each of the
positive and negative MI groups were added as between subject factor. Thus two separate
mixed factorial ANOVAs (4 × 2 × 2) were computed for valence ratings obtained from
four mood techniques (MT) for happy and neutral mood conditions (MC) in two blocks
(B) of both Positive and negative MI groups. Similarly 2 mixed factorial ANOVAs were
computed with arousal ratings. Succeeding paired sample t-test provided comparison
amongst four mood techniques for efficacy in inducing anticipated mood state.
Selection of video songs, instrumental music and pictures was based on the mean scores.
228
Table 4.1a
95% CI
Mood Songs
M SD LL UL
Happy
Sad
The table showed mean scores of ratings for happy and sad video songs. Bum
Bum Bolay was selected from the happy list as it showed the highest mean score (M =
8.53, SD = 1.18). Song 1 (M = 1.66, SD = .89) and song 2 (M = 1.66, SD = .61) were
229
equally rated as the most sad songs, however Maa (song 1) was selected being more
Table 4.2a
95% CI
Mood Instrumental
M SD LL UL
Happy
Sad
Table 4.2a demonstrates the means and standard deviations for ratings of happy
and sad instrumental music within 95% confidence interval. Sound track of Doraemon
230
and PTV sad instrumental music were selected as being most significantly effective in
inducing happy (M = 8.86, SD = .99) and sad mood (M = 1.73, SD = .79) respectively.
music was based on the highest rating obtained, and the sad music was chosen on the
basis of lowest mean scores. The neutrality of the pre-selected instrumental music piece
was validated by all the children (n = 30) who rated it on the same eleven point Likert
scale. The musical excerpt was rated near 5 ranging from 4 to 6 points.
231
Table 4.3a
Table 4.4a
Table 4.5a
Results of Table 4.3a indicated that pictures having mean score of less than 3
were categorized as sad. However three pictures that obtained a zero rating by more than
3 judges were excluded from the final set being extremely traumatic and thus sensitive to
be used with children. Thirty pictures having lowest mean scores were selected for the
final set. According to the results in Table 4.4a, initially, pictures having a mean score of
greater than 7 were identified as happy. Yet the final set included 30 pictures with highest
mean scores and all of them are greater than 7.5. Table 4.5a presents the neutrality of the
pictures that was judged on the basis of mean scores closer to the central point of the
rating scale (4.8 to 5.2). However 30 pictures mostly ranging from 4.9 to 5.1 comprised
Based on the above mentioned criteria, three set of pictures were finalized: happy
(30), sad (30) and neutral (30) (see App. F for sample pictures).
The data were analyzed to examine changes in arousal and valence as a function
of watching video song, listening instrumental music, watching pictures alone and in
combination with instrumental music for happy, sad and neutral mood conditions.
Negative group participants listened to sad and neutral music, and positive group
Table 4.6a
Variables NA NB PA PB (N = 64)
(n = 16) (n = 16) (n = 16) (n = 16)
BLA
Gender (%)
NA: Negative Block A; NB: Negative Block B; PA: Positive Block A; PB: Positive Block B. BLM: Baseline Mood. BLA:
Baseline Arousal
watched the video song, followed by picture presentation and combined MIP. For
interchanged to ascertain the counterbalancing. MIP groups did not differ in age, F (3,
63) = 0.85, p = ns and baseline mood, F (3, 63) = 0.67, p = .ns. Gender distribution was
also equal across groups (50%). Age differences within gender groups were also non-
Valence and arousal of techniques across MI groups and gender. Tables 4.7a
and 4.8a present the descriptive statistics for valence and arousal in positive and negative
MI groups.
Table 4.7a
Descriptive Statistics for Mood Induction Trials within Positive MI Group (n = 32)
Skewness Kurtosis
Trials Min Max M SD Statistic S.E Statistic S.E
BLM 3 7 4.968 .739 .050 .414 2.041 .809
BLM = Baseline Mood, HV = Happy Valence, NV = Neutral Valence, Inst = Instrumental, Vid = Video,
Baseline mood rating of the positive MI group showed neutral mood tendency (M
= 4.96, SD = .73). Happy mood was more effectively induced by video song (M = 2.039,
Neutral mood was most successfully induced through picture presentation (M = 4.90, SD
= .58). Kurtosis and skew values of valence for mood stimuli were within acceptable
ranges except for happy video song, set of happy pictures and neutral combination
237
techniques that showed minor increases from typical standard suggested (3.00 for
Table 4.8a
Descriptive Statistics for Mood Induction Trials within Negative MI Group (n = 32)
Skewness Kurtosis
Trials Min Max M SD Statistic S.E Statistic S.E
BLM 3 7 5.031 .822 -.431 .414 1.628 .809
BLM = Baseline Mood, SV = Sad Valence, NV = Neutral Valence, Inst = Instrumental, Vid = Video, Pic =
Mean scores indicated that baseline mood of negative MI group was neutral (M =
5.03, SD = .82). Highest valence scores were obtained for video song (M = 8.03, SD =
.64) and combination MIP (M = 8, SD = .62) which indicated that both of these
techniques were more successful in inducing sad mood. All the neutral MI techniques
depicted valence scores near the central point (5) which shows the neutrality of the
stimuli.
238
The data were analyzed for differences in mean valence and arousal across two
Table 4.9a
Source DV SS df MS F η2
negative MI groups in ratings of valence for instrumental music, F (1, 60) = 804.05, p <
239
.001, η2 = .931; video song, F (1, 60) = 1172.3, p < .001, η2 = .951; pictures, F (1, 60) =
1067.3, p < .001, η2 = .947, and combination, F (1, 60) = 1748.3, p < .001, η2 = .967
techniques. Boys and girls perceived the stimuli in similar manner as no significant
gender differences were found for valence of most of the techniques overall and as a
function of positive or negative group. The only significant difference across gender was
observed for combination technique, F (1, 60) = 388.2, p < .05, η2 = .061.
Figure 4.2a
Figure 4.2a shows clear differences in valence of all mood induction techniques
across positive and negative mood induction groups in an expected direction. Mood
Manikin (SAM) indication happier mood while negative MI group rated the stimuli
Figure 4.3a
Table 4.10a
Source DV SS df MS F η2
differed in level of arousal. Positive and negative groups significantly differed in arousal
ratings of instrumental music, F (1, 60) = 434.56, p < .001, η2 = .879; video song, F (1,
60) = 717.55, p < .001, η2 = .923; pictures, F (1, 60) = 311.91, p < .001, η2 = .839 and
combination technique, F (1, 60) = 968.13, p < .05, η2 = .943 on SAM. Main effect of
gender on arousal level and its interaction with MI groups was found insignificant.
Figure 4.4a
Differences in level of arousal across positive and negative groups for mood
induction techniques are apparent (Figure 4.4a). It indicates high arousal (low scores) for
Figure 4.5a
Figure 4.5a depicts that level of arousal for positive and negative MI techniques
by observing the extent to which the valence and arousal scores changed across baseline
mood and 7 subsequent trials which include neutral materials serving as pre-MI and
subjects effect of mood techniques (4) and mood types (2); and between subject effects of
blocks within Positive and Negative MI groups for valence and arousal. It followed
paired t-test to identify that which of the techniques produced more desirable outcomes.
245
Table 4.11a
Source SS MS df F p η2
Mauchly’s Test of Sphericity showed that the sphericity proposition had not been
violated, χ2 (2) = 0.701, p = ns, and therefore, all statistics are reported with sphericity
assumed.
246
conducted with MI techniques (4; instrumental, video, pictures, combination) and mood
conditions (2; happy, neutral) as within subject factors, and blocks (2; instrumental MI
first vs. picture MI first) of randomized trials as between subject factor. Initial baseline
mood and 7 subsequent trials alternating for happy and neutral moods were entered as
within subject factors resulting from combination of 4 × 2. Significant main effects have
been found for four type of mood induction techniques, F (1, 30) = 8.035, p < .001, η2=
.211. Valence of happy and neutral MIP also differed significantly within subjects, F (1,
30) = 1166.3, p < .001, η2= .974, which indicates the efficacy of MIPs in producing
desirable affect. None of the interactions could yield significant effects. Results of
between subject effects also indicated no significant difference in valence scores across
Figure 4.6a
Figures shows that first (Instrumental music) and third (Picture only) mood
techniques had higher mean scores on Self Assessment Manikin as compared to second
(Video song) and fourth (Combination) technique. Lower scores on SAM refer to higher
positive valence, whereas higher scores indicate negative valence. Happy (Type 2) and
neutral (Type 1) mood conditions seem to vary in valence as neutral mood valence is
ranging from 4.5 to 5, and happy mood valence is clustered around 2-2.5. Neutral mood
condition was assessed initially as baseline mood (1), then mood induced by instrumental
music (2), picture presentation (3) and the combination technique (4). Neutral condition
did not involve video based mood induction. Both mood conditions illustrate the similar
Table 4.12a
95% CI Cohens’
Pairs df t p LL UL d
Subsequent pair wise comparisons were performed through paired sample t-test.
Analysis revealed significant differences in instrumental and video based happy mood
valence scores, t (31) = 2.10, p < .05, as video song induced more happiness (M = 2.09,
(M = 2.43, SD = .564) and instrumental excerpt did not differ in valence and induced
happy mood to the same extent, t (31) = .27, p = ns. Instrumental music significantly
differed with combination method (M = 2.03, SD = .537) in inducing happy mood, t (31)
= 2.94, p < .01. Emotional pictures presented alone were found less effective in
suggesting happy mood state as compared to video song, t (31) = -2.24, p < .05, and
combination method was found more effective than pictures, t (31) = 3.04, p < .01. Video
song and combined MIP did not differ significantly in inducing happy mood state.
249
positive MI group.
Table 4.13a
Source SS MS df F p η2
Mauchly’s Test of Sphericity indicated that the assumption of sphericity had not
been violated, χ2 (2) = 0.701, p = ns, and therefore, all statistics are reported with
sphericity assumed.
250
Four types of happy mood techniques were significantly different in arousing the
desired mood, F (1, 30) = 49.13, p < .001, η2= .621. Similarly level of arousal was
significantly different across happy and neutral mood conditions, F (1, 30) = .666, p = ns,
η2= .022. These results suggested that happy MI techniques produced higher arousal level
as compared to neutral techniques (see Figure 4.6). Different mood techniques resulted in
arousal differences across happy and neutral types of MIP, F (1, 30) = 66.77, p < .001,
η2= .690. Both of the randomized blocks (Instrumental MI first vs. picture MI first)
within negative MI group did not differ significantly in arousal level, F (1, 30) = .25, p =
Figure 4.7a
Figure 4.7a shows that level of arousal is different across happy and neutral mood
conditions, yet the magnitude of difference is less clear for instrumental music technique.
252
Figure 4.8a
Happy and Neutral Arousal for Mood Techniques across Negative MI Blocks
Both of the blocks (Instrumental music first vs. picture first) were alike in
Table 4.14a
95% CI
Pairs df t p LL UL
Four types of mood induction techniques did not significantly differ with each
computed for evaluating the techniques and types of mood in blocks of negative MI
group.
254
Table 4.15a
Source SS MS df F p η2
Mauchly’s Test of Sphericity indicated that the assumption of sphericity had not
been violated, χ2 (2) = 0.701, p = ns, and therefore, all statistics are reported with
sphericity assumed.
255
Level of arousal was different across techniques F (1, 30) = 9.09, p < .001, η2=
.233. Arousal differences were also found significant for sad and happy mood conditions
F (1, 30) = 1365.3, p < .001, η2= .978 and the interaction between mood techniques and
mood conditions F (1, 30) = 5.624, p < .001, η2= .158. Negative MI blocks yielded non
Figure 4.9a
Mean Valence Scores of Four MI Techniques across Blocks within Negative MI Group
Figure 4.10a
Mean valence scores of four MI techniques were higher for sad condition than
lower scores of neutral condition. It demonstrates that sad materials induced sad mood
effectively and neutral materials were also successful in suggesting neutral mood. Second
(Video song) and fourth (Combination) were more effective as compared to first
Table 4.16a
95% CI Cohens’
Pairs df t p LL UL d
change across four types of MI techniques were computed. Instrumental music and
Emotive Picture MIPs, t (31) = .780, p = ns, as well as the Video song and combination
MIPs, t (31) = .226, p = ns, partial, also yielded non-significant results. All the other pairs
Table 4.17a
Source SS MS df F p η2
participants in significantly different manner, F (3, 90) = 72.81, p < .001, η2= .708.
Similarly main effects of mood conditions, F (1, 30) = 5.624, p < .001, η2= .158 and
259
interaction of mood condition and techniques, F (1, 30) = 5.624, p < .001, η2= .158 was
also significant.
Between subjects effects showed that the two blocks were not significantly
different in arousal levels as a result of mood induction, F (1, 30) = .073, p = .789, η2=
.002.
Figure 4.11a
Level of Arousal for Sad and Happy Mood Conditions across MI Techniques
260
Estimated marginal means of arousal across sad and neutral condition were
different for instrumental music and picture only techniques. However video song and
combination techniques were less arousing across both sad and neutral conditions.
Figure 4.12a
Mean Arousal Scores of Four MI Techniques across Blocks within Negative MI Group
Arousal levels for four MI techniques were the same in both blocks of negative MI group.
261
Table 4.18a
95% CI
Pairs df t p LL UL
Subsequent t-test paired comparisons that Video song, t (31) = .780, p = ns,
Picture only, t (31) = .780, p = ns and Combination, t (31) = .780, p = ns techniques were
more arousing than instrumental music. Pictures alone and in combination with
background music along with video song did not differ in arousing mood level of
To sum up, analysis of the scores on the mood valence and arousal rating scales
showed that the proposed effect of the all the four MIPs was obtained across happy, sad
and neutral conditions However, video song and combination of pictures and background
instrumental music were more effective than instrumental only and picture only
techniques
262
Discussion
(instrumental music, video song, emotive pictures alone and emotive pictures combined
with background instrumental music) in order to determine their efficacy for inducing
happy, sad and neutral mood states in elementary school children. Mood suggestive
materials for each of the four techniques were first selected and rated by independently
Studies using music mood induction procedures (MMPI) have mostly relied on
classical instrumental music excerpts that had been producing robust induction effects on
mood states of people (Vastfjall, 2002). Similarly emotion laden photographs also have
powerful evidence in altering mood state (Goritz, 2006; 2007). However most of such
studies are conducted in Western countries and the materials are selected from
standardized data bases that may not fit to Pakistani cultural context. Therefore music
(instrumental and video) pieces were selected mostly from familiar and famous film
content of Pakistan and India as music across borders is equally understandable and
popular due to language comprehension and the use of same genres and instruments. To
Limited number of songs and instrumental music was selected to avoid ambiguity for
child raters. Hunter et al. (2008) have followed same kind of approach with child sample
by using only five musical stimuli for particular affect category rather than presenting a
long list as often done with adult samples. Following the criteria used by several
excerpt with highest mean rating was selected for further evaluation. Hence we relied on
mood scores only that were provided by two groups of children who independently
263
listened to the list of each type of music. Highest mean scores were obtained for Bum
Bum Bolay (TZP, 2007) from list of happy songs. The music of this song follows a high
arousing cheerful tempo, and the videos portrays children singing and dancing in a
wonderful blend of joy and excitement that made children feeling happier while listening
were fond of high and fast tempo music (Montgomery, 1996). From the list of sad songs,
Maa and Janay Kub Hongay Kum were equally rated with lowest scores (indicating
greater sadness) than other three songs. Maa was selected being more comparable to
happy song (Bum Bum Bole), as it is from the same movie (TZP, 2007) and depicts a
child being sent to a boarding school where he missed his family especially mother. The
findings are in line with the previous studies that reported of children s’ preference for
age appropriate and popular music (Brittin, 2000; Demorest & Schultz, 2004). None of
video songs was considered to be neutral in music selection phase of the study as song’s
As far as instrumental music is concerned, movie and television sound tracks have
been claimed to be more popular and preferred form of music by children in a study by
Roulstone (2006). We therefore opted for soundtracks of popular cartoon series and
movie songs, and a few playbacks of television for eliciting sad and happy mood. A piece
of instrumental music employing a blend of high and low tempo was selected as the only
neutral excerpt. Judges in selection phase considered it boring and not arousing. Sound
track of Doraemon (the most famous cartoon series) evoked more happiness and an
unknown musical piece (usually played back of tragedy clips on Pakistani television
264
channels) that was downloaded from internet was regarded as the saddest instrumental
excerpt. The aspect of familiarity was evident in ratings which confirms that age and
Mood suggestive pictures were rated by judges for sad, happy and neutral affect.
Here too, top rated pictures comprised happy set, which mostly included pictures of
happy children in different settings such as with family, friends or at play. It indicates
that pleasant and intimate social interactions are considered as source of happiness.
Cheerful pictures of infants were also rated high for evoking happiness. Pictures of
nature’s beauty, celebrations and festivals were also rated high on positive affect. The
lowest rated pictures comprised sad set of pictures. Pictures of illness, death, destruction,
accidents, misery and poverty portrayed with intensity were rated high on negative affect.
Pictures rated closer to the center point were categorized as neutral. These pictures
illustrated unemotional content such as a busy or empty road, building, furniture, routine
life of people. Mood induction studies have widely used pictures for mood priming, and
International Affective Picture System (IAPS, Lang et al., 1997) is the most extensively
used standardized set of pictures. It is also based on similar emotional themes as selected
in our study, yet the content is based on virtual reality paradigm. Yet we preferred to
collect culture relevant and age apt pictures capturing natural and original scenarios.
All the four techniques were evaluated for effectiveness in inducing sad, happy
blocks for positive and negative mood induction groups. Results assured that participants
of both groups did not differ significantly from each other before the mood manipulations
265
commenced. The experiment was designed carefully so that it could be easily replicated
by other studies.
Findings confirm that the techniques overall were effective in altering mood states
in desired manner. Subsequently, the relative efficacy of the four techniques was
observed with a separate analysis of valence and arousal. For positive MI group,
watching and listening to video song and viewing evocative pictures while listening to
music were found more powerful mediums in inducing both happy and sad moods. A
simple explanation is that watching and listening combined can attract more attention as
compared to only listening or only watching the content especially for children.
Therefore both video song and picture music combo evoked pleasant affect and were
highly arousing in positive MI group. Whereas sad video song and picture combined with
music elicited unpleasant affect and low levels of arousal. Extant of evidence suggest that
materials perceived to express happiness induce higher arousal and positive affect, while
those depicting sad impressions induce states of lower arousal and negative affect (e.g.
Schellenberg, 2012).
Neutral content of pictures was stronger in evoking proposed mood than neutral
piece of music. It produced lower arousal and affect. Valence and arousal both are
important dimensions of a specific mood. Previous section of rating materials missed the
dimension of arousal as Self Assessment Manikin was not available at that time and it
was difficult for children to rate the arousal dimension with verbal label. Therefore later
which shows that experimental design yielded fair results. The result for mood induction
266
is comparable to findings in the existing literature. Brenner (2000) suggested that video-
mediated affect inductions were more advantageous as compared to the other techniques
due to the ease of administration and more enduring effects to mood states. Both of the
techniques found most effective in the current study could relate to this type method.
Music enhances the effect of image-base stimuli, either static or moving. Picture-based
induction materials have been combined with music in different studies to study the
impact of mood changes on health-related behaviors (Conklin & Perkins, 2005; Perkins
et al., 2008) and emotion perception (Baumgartner et al., 2006). Although movie songs or
video songs other than movies have not been directly studied with reference to mood
induction especially with children however a combination of film and music had been
used to study the influence of mood on social appraisals and categorization (Halberstadt
The study contributed to the extant literature on mood induction and was first of
its kind in South Asian region especially Pakistan. Based on the findings of present study
both video song and emotive pictures combined with music can be used as an effective
method of mood induction with children. Further investigation is needed to evaluate the
music characteristics and appraisal of each picture included in MI set for valence and
arousal. Conducting the experiment in laboratory under more controlled conditions and
important source of information for observers (Van Kleef et al., 2009), whereas mood
offers a subjective source of information (Schwarz & Clore, 1983). This information
might have an impact on reactions as people in different mood states derive different
meanings by observing facial expressions. So the student’s mood state and their appraisal
2004). Hence the emotion recognition ability could also be linked to quality of friendship
The present study mainly addressed the core question of the current endeavor that
victimization (Crick & Dodge, 1994; Sutton et al., 1999). Attentional bias as result of
mood congruence has also been explored. Additionally, possible links between emotion
recognition ability and quality of friendship in bully/victim groups were also explored.
The sample for the current study was a subset of participants identified already as bullies,
The study was divided into two sections. The first section was based on
Based on the findings of study 2, two mood induction techniques were utilized for
priming the participants with sad, happy and neutral mood states. Those were video song
268
and emotive pictures with background music. Both of these methods had demonstrated
relationship between emotion recognition ability and the quality of friendship among
of bully/victim groups were also evaluated. For this purpose, data from study 3 (section
uninvolved students.
3. To examine whether happy and sad moods boost or hinder the recognition of mood
Hypotheses
1. Bully group will perform better than victims and bully-victims, on emotion
recognition task.
269
2. Boys and girls will differ in their ability to recognize facial expressions of
emotions.
congruent direction.
4. Mood congruency effect would differ across bullies, victims and bully-victims.
Method
Participants
Participants were 394 students (M = 10.77, SD = 1.07; 45.7 % boys) from 4th to
6th grade in 15 schools of Lahore. All 9 to 12 years old students who participated in the
third phase were invited to take part in this experimental phase. Authorities of three
schools refused to participate further. Only 470 showed their willingness to participate.
Thus an informed parental consent was obtained from these 470 available students. We
received 46 refusals. Thus the experimental phase was scheduled with 424 students from
9 schools. Fourteen of these 424 participants were absent on the day of experimental task
task, while three participants reported discomfort during mood induction procedure and
left the study. Three participants who did not display mood that was consistent with their
assigned condition were also excluded from the final data set. Pre-induction valence
ratings of 2 participants in the picture with music condition and 1 in video-based music
270
condition were outside of the normal range (three standard deviations) and therefore were
removed. Four participants completed the experimental procedures but their data was lost
Measures
Two experimental tasks and a manipulation check pre and post mood induction
1. Mood state assessment. Children by the age of 4 to 5 years are able to report on
their basic emotions correctly (Luby et al., 2007). Single question self-reports offer the
most accurate account of the emotions experienced by children and being the least liable
to demand characteristics have been strongly recommended for child samples (Brenner,
2000). Therefore, the Ottawa-Georgia Mood Scales (Cheng & Ward, 2005) for children
and youth were considered to evaluate mood states before and after manipulation. It
feelings around mood (from sad to happy), anxiety (from confident to scared), and anger
(from calm to angry). I have used one of the scales to rate happy, sad and neutral mood
2. Mood induction techniques. As both vi.deo song and emotive pictures combined
with background music demonstrated highest efficacy, we decided to use both of these
techniques successively to present sample in a counterbalanced order for happy, sad and
neutral conditions. Thus half of the participants in each condition were supposed to watch
the video first and then take the picture/music task. Other half first experienced
picture/music followed by video song. The decision was made in order to induce more
271
enduring mood state as participants had to undergo the emotion recognition task that
combination would facilitate the participants to maintain the mood state with which they
Happy song (Bum Bum Bolay from the movie Tare Zameen Par) was 5 minutes long.
Sad song (Maa from Tare Zameen Par) was also of similar length. Three sets of 30
pictures for each mood condition were presented through Microsoft power point along
with mood congruent back ground music. Picture/music task was approximately three
background music was also adjusted accordingly after editing. Thus MIP lasted for 8-9
minutes approximately.
was measured with the help of Facially Expressed Emotion Labeling Test (Kessler,
Bayerl, Deighton, & Traue, 2002). The test has manifested good internal consistency
having Cronbach’s alpha approximating .77. It has been translated into Urdu (Malik,
Khawar, Chaudhry & Humphreys, 2010) and also validated for Pakistani population
across different age groups including children (Khawar, Malik, Maqsood, Yasmin &
Habib, 2012); for which it proved to be a reliable measure (Cronbach’s alpha, α = .74) for
the current study. This is a computer program in which color photographs of neutral faces
followed by the same faces expressing a certain basic emotion appears on the screen, and
subjects have to identify the adequate emotion shown. Response options are presented in
forced choice paradigm. The pictures used by the FEEL test are taken from JACFEE
(Japanese and Caucasian Facial Expression of Emotion) series (Matsumoto & Ekman,
272
1988). The six displayed basic emotions include: anger, sadness, fear, disgust, happiness,
and surprise. The task consists of 42 pictures (7 pictures for emotion) that are presented
on a computer screen. Thus the scores based on response accuracy ranges from 0 to 42.
Reaction time is also measures in terms of average time a subject takes to recognize an
emotional expression. Presentation of the stimulus and selection of the emotion scene
Procedure
First of all an informed consent form was distributed amongst the students
previously assessed for OBVQ and FQS in third phase. After enlisting all the students
who provided the written parental consent for this study; date, time and location of the
experiment was scheduled with the help of school administration. The experimental study
was conducted in computer labs of the respective schools to ensure the controlled and
consistent conditions. On specific day scheduled for the experiment, the participants were
first assessed for their current mood state using Ottawa Geogia Mood scales. Bullies,
victims, bully-victims and uninvolved students were then randomly assigned to one of the
three mood induction conditions (happy, sad and neutral). As we decided to use both
video songs and pictures combined with background music being presented successively,
the order of the presentation was counterbalanced across bully/victim groups. All the
participants in neutral condition received the same stimuli that were set of neutral
pictures with background music. The experiment was conducted individually. Before
staring the experiment, two practice trials were used to familiarize the child to the laptop.
Subjects opened my computer from desktop and moved to the desired folder by
him/herself.
273
Figure 4.13a
Study 3 (1)
(N = 817)
Sample
Study 4
(N = 394)
The participants were first assessed for current mood state with the help of Ottawa
Georgia Mood Scale. After getting primed with happy, sad, or neutral moods they were
again assessed for mood state. They had to indicate on a scale (from sad passing through
neutral to happy) how sad, or happy it made them feel. Then they were briefly instructed
to perform emotion recognition task. All subjects could read and understand the six labels
presented after 500 ms of the stimulus picture. For each one of the six emotions, eight
pictures were displayed. In a warm-up round the subjects saw six example faces (one for
each emotion) and were asked to indicate the displayed emotion by pressing the
274
appropriate computer key. They got feedback on whether their decisions were correct or
incorrect. This warm-up was designed to make the subjects familiar with the testing
procedure. There remain 42 pictures for the main test (six emotions with seven pictures
each) which were presented in the same manner without providing any feedback. First a
picture of a neutral face was displayed on the computer screen for 1.5 seconds with a
short beep to attract the attention of the subject. Then there is one-second break during
which the screen is gray. After that short break the stimulus itself is presented on the
screen for exactly 300 ms. The stimulus is a picture of the same person as seen before
with a neutral face this time displaying one of the six basic emotions. First showing the
neutral and then the emotional face imitates natural conditions where the emotion often
undergo happy mood induction in order to repair the negative affect. Twenty two
participants availed this option. All of them were thanked for their cooperation.
275
Figure 4.14a
Figure 4.15a
Results
Results of the final phase, (Phase IV) reveal the experimental outcome evaluated
with the help of independent samples and repeated measures analysis of variance Main
effects, interactions and effect sizes were calculated for main variables. Post hoc tests
Table 4.19a
Gender
Victim 36 46 82
Uninvolved 46 61 107
Table 4.19a shows that bully/victim groups did not differ in terms of gender
Table 4.20a
Mood Conditions
Victim 22 33 27 82
Uninvolved 42 34 31 107
Table 4.20a shows that bully/victim groups were equally distributed to three
Descriptive Statistics
information that could further guide the analyses. The distribution of variables was
Table 4.21a
Table 4.21a shows the mean accuracy scores for the FEEL and individual
emotions. Highest accuracy was found for happy facial expressions (M = 6.54, SD = .96).
Therefore a few cases that scored lower on this emotion category were identified as
extremes. Facial expression of fear was least accurately recognized by the total sample
(M = 2.86, SD = 1.94). Score on overall emotion recognition ability was also not so high
Table 4.22a
Skewness Kurtosis
Valid N 394
The data on FEEL test were approximately normally distributed except for happy
emotion recognition as most of the participants scored high on this category. Kline (1998,
2005) has suggested that skewness value below 3.0 and kurtosis below an absolute value
Manipulation Check
Two way ANOVA with three mood induction condition (sad, happy and neutral)
as the independent variable and participants’ pre and post mood ratings as the within
subjects dependent variable was employed to test if mood induction worked adequately.
SPSS does not provide sphericity estimates for within subject variable having less than
three levels. Greenhouse-Geisser value in such cases is mostly 1 (Field, 2013). Similar
results were obtained for the present study suggesting that assumption of sphericity had
been met.
Table 4.23a
across Conditions
Sources SS df MS F η2
Within-Subjects
Between-Subjects
**p < .01, ***p < .001; MR = Mood Ratings, MC = Mood Conditions
282
The main effect of mood as within subject variable was significant, F (1, 391) =
6.047, p <.014, η2 = .015, which indicated that all the subjects significantly differed
between their pre and post mood ratings. Interaction effect of mood and conditions was
also significant, F (2, 391) = 988.357, p <.001, η2 = .835. Analyses showed that
participants in all the three mood conditions significantly differed in their pre and post
mood ratings. Participants felt significantly happier after going through happy mood
induction (M.diff = -3.90), t (132) = -27.34, p < .001; sad after sad mood induction
(M.diff = 4.44), t (131) = 31.82, p < .001; and that neutral mood induction successfully
Between subjects mood effect were also found significant, F (2, 391) = 2135.007,
p < .001, η2 = .787 which indicated that participants in each mood condition (sad, happy,
Figure 4.16a
Pre and Post Induction Mood Ratings across three Mood Conditions
The Figure (4.16a) clearly shows the elevated scores after happy mood induction,
decreased trend after sad mood induction and a stable pattern after neutral mood
induction. It demonstrates that all the mood induction procedures were effective in
ability to recognize facial expressions of emotions. Firstly, FEEL total score was
analyzed for possible, gender and Bully/victim group differences using Univariate
ANOVA.
Table 4.24a
Source SS df MS F p η2
Bully/victim group showed significant main effects on FEEL total score, F (3,
386) = 5.86, p < .001, η2 = .044. Gender could not yield any significant effect on overall
emotion recognition ability, F (1, 386) = .248, p = ns, η2 = .001. However, significant
differences were found for accuracy responses on FEEL as a result of gender and
Figure 4.17 a
7.12). Further, emotion recognition ability of bully-victim group was poorer as compared
286
to victims and bullies. Bullies seemed to perform relatively better on FEEL test. These
Figure 4.18 a
Mean Scores of Boys and Girls on FEEL Test across Bully/victim Groups
The figure demonstrates clear gender differences in victim group. Bullies and
Table 4.25a
95% CI
Table 4.25a shows that students classified as bullies, victims and bully-victims
did not differ in recognizing facial expression of emotions, yet these three actively
groups, Mood condition and Gender) and one within-subjects variable (six different facial
any inconsistencies in processing any specific facial expressions of emotion. Results are
Table 4.26a
Source SS df MS F p η2
Within-Subjects
Between-Subjects
Analysis of FEEL data for 6 emotion types revealed significant main effect for
Bully/Victim groups as between subject factor, F (3, 370) = 5.97, p < .001, η2 = .046, and
emotions, F (5, 1850) = 191.52, p < .001, η2 = .341, as within subject factor. The
interaction between Bully/Victim groups and accuracy score on types of emotions was
All the other main effects for variables (Mood Condition and Gender) and their
interaction as a function of both within and between subjects were found insignificant.
However, for between groups, interaction of bully victim groups and mood conditions
indicated some differences though not found significant. The data were further evaluated
using MANOVA for possible main effects and interaction of mood conditions across
bully/victim groups.
291
Table 4.27a
Dependent
Source Variable SS df MS F p η2
MC Anger 2.128 2 1.064 .246 .782 .001
Sadness 2.402 2 1.201 .295 .745 .002
Disgust .025 2 .012 .003 .997 .000
Surprise 1.135 2 .568 .119 .888 .001
Happiness .408 2 .204 .219 .804 .001
Fear .505 2 .252 .068 .935 .000
FEEL 4.968 2 2.484 .047 .954 .000
BVS Anger 31.358 3 10.453 2.415 .066 .019
Sadness 42.037 3 14.012 3.443 .017 .026
Disgust 60.927 3 20.309 4.274 .006 .032
Surprise 110.073 3 36.691 7.670 .001 .057
Happiness 1.975 3 .658 .706 .549 .006
Fear 26.243 3 8.748 2.340 .073 .018
FEEL 1024.500 3 341.500 6.415 .001 .048
MC × BVS Anger 80.599 6 13.433 3.104 .006 .046
Sadness 17.167 6 2.861 .703 .647 .011
Disgust 29.564 6 4.927 1.037 .401 .016
Surprise 40.711 6 6.785 1.418 .206 .022
Happiness 2.901 6 .484 .519 .794 .008
Fear 27.446 6 4.574 1.224 .293 .019
FEEL 682.830 6 113.805 2.138 .048 .032
Error Anger 1653.372 382 4.328
Sadness 1554.723 382 4.070
Disgust 1815.366 382 4.752
Surprise 1827.478 382 4.784
Happiness 356.217 382 .933
Fear 1428.110 382 3.739
FEEL 20334.362 382 53.231
MC = Mood Condition; BVG = Bully/Victim Groups
292
Main effects of mood condition were not found significant. Participants in three
mood conditions (happy, sad and neutral) equally recognized all the emotions showing no
congruence effect of mood. Recognition of three basic emotions and total FEEL score
was significantly different across bully/victim groups except for anger recognition.
Significant main effects were found for sadness, F (3, 382) = 3.44, p < .01, η2 = .026;
disgust, F (3, 382) = 4.27, p < .01, η2 = .032; and surprise, F (3, 382) = 7.67, p < .001, η2
= .057. Bully/victims group did not differ in recognizing anger, happiness and fear.
Interaction of mood conditions and bullying involvement was also found insignificant for
emotion categories except for anger, F (1, 58) = 8.33, p < .001, η2 = .046. Interaction
effect was also significant for total FEEL score, F (1, 58) = 8.33, p < .001, η2 = .032.
Bully/victim groups were further evaluated using LSD Post Hoc comparisons. Interaction
Table 4.28a
Post Hoc Comparison among Bully/Victim Groups for Individual Emotions on FEEL
95% CI
Significant mean differences were found for anger recognition among bullies,
bully-victims and uninvolved students. Recognition accuracy was higher for uninvolved
294
students as compared to bullies and bully-victims. Victims and uninvolved students were
alike in recognizing anger expressions. Bully-victim group showed poorer recognition for
for disgust was higher for uninvolved group as compared to the other three groups
poorer recognition ability for surprise as compared to uninvolved group. Fear was better
Figure 4.19a
accurately as compared to those who underwent sad or happy mood induction. Lowest
Figure 4.20a
performed better than victims in happy and sad mood states. Contrarily, uninvolved
group in neutral mood condition scored lower on FEEL test as compared to participants
run with the following two independent variables: Involvement in bullying (as bully,
victim, bully/victim and uninvolved) and Mood conditions (happy, sad and neutral). All
the emotions and total FEEL score were the dependent variables. Participants’ full score
on Colored Progressive Matrices (Raven’s CPM) was included as a covariate, due to the
Table 4.29a
ANOVA showing Differences in Reaction Time across Bully/Victim Groups and Mood
Conditions
Source SS df MS F p η2
Only main effect of Bully/Victim group was found significant, F (3, 381) = 3.9, p
< .01, η2= .03. Mood induction and bully/victim groups in different mood condition
Table 4.30a
Post Hoc Comparisons among Bully/Victim Groups for Reaction Time on FEEL
95% CI
LSD Post Hoc comparisons among bully/victim groups revealed that students
uninvolved in bullying significantly took lesser time in completing the FEEL tests as
compared to bullies (M.diff = -68, p < .01), victims (M.diff = -63, p < .01) and bully-
Figure 4.21a
uninvolved group responded quickly to the stimuli as compared to the other three groups.
Discussion
The study was conducted to check the assumption based on the argument between
social cognitive model of Sutton et al (1999) and social skill deficit model offered by
Crick and Dodge (1994) if bullies would have better emotion recognition ability as
Lemerise and Arsenio’s (2001) account for the role of emotions in SIP, we empirically
evaluated if mood would affect the emotion recognition skill in a congruent manner.
Boys and girls were also expected to differ in recognizing facial expressions of emotions.
The above mentioned postulations were not mainly supported by the findings of
present study. Though the results suggested significant differences among bully/victim
groups for emotion recognition accuracy, however bullies were equally deficient in
decoding facial expressions of emotions as victims and bully-victim groups. These three
groups performed equally worse on FEEL test, supporting an overall deficit in relation to
bullying, a finding that was found consistent with SIP literature and aggression (Orobio
de Castro et al., 2003). The results did not favor the notion offered by Sutton et al. (1999)
that competent social cognitive skills could cause incompetent behaviors, rather they are
consistent with Crick and Dodge’s (1999) connotations. Uninvolved group showed
significantly higher accuracy scores than other three groups. Though not significant yet
the bully-victim group was found to be less accurate in discerning emotional expressions
as compared to those who were victimized or bullied others. These results should be
interpreted with results for reaction time. FEEL measures average reaction time for
recognizing each emotion and findings compliment the previous result of our study.
Uninvolved group took significant lesser time to choose the response option as compared
301
with groups directly involved in bullying. Students who bully others consumed slightly
more time than other two groups, yet the difference was not found significant.
observed in the present research. Perhaps, these differences emerge later in life as
accuracy is concerned, happy and sad expression were recognized better than others. Fear
followed by anger was least accurately recognized emotions by total sample. This general
deficit in recognition of anger and fear could also be explained by younger age of the
participants. It can be suggested that this inaccuracy in recognition of two basic threat
related cues contribute to maladaptive peer relationships and behavior patterns (Eisenberg
& Fabes, 1992). The ability to accurately recognize, evaluate and appreciate the
emotional expressions of others has been considered a basic prerequisite for good social
interactions (Saarni, 1999). The results are in line with those found by Woods et al.
(2009) who also reported least accuracy in recognizing angry expression. They also
found higher mean errors in recognizing angry and fearful expression by those who had
been bullied either by relational or physical means. This inability to identify fear and
anger leads children to misinterpret situational cues that could facilitate in aggression
inhibition (Blair, 1999). The findings could be further associated with the type of
Results showed that mood was effectively induced and this adds to the reliability
of the techniques developed in the first phase of this study. However, mood congruence
could not be observed in general. Overall effect of mood on emotion recognition was
302
found non-significant however interaction effect of mood and bully/victim groups led to
the conclusion that victimized children under neutral mood condition used to recognized
the angry facial expressions better than when they were in sad or happy mood states. This
was also found true with general ability to interpret emotional expression (total FEEL
scores). Thus first three assumptions were not accorded. However partial support was
Therefore the present results are in accordance with the view that victimized
thus facilitated access to representations of, anger when they were not influenced by pre-
existing mood. This ability to use partial information about a facial expression of anger
circumstances where the incidence of threat and the chances of injury are elevated
(Woods et al., 2009). The finding is further supported by another result of the present
study, as uninvolved group was significantly found more accurate in recognizing anger as
compared with bully and bully-victim group. No significant differences were found
between uninvolved and victimized children. The results are somewhat consistent with
previous studies in explaining that victims are not deficient in recognizing all kinds of
expressions (Camodeca et al., 2003). However it remains a question as why victims could
not take benefit of this better recognition and avoid the conflicting situation.
Bullies and bully-victim were also less accurate in identifying sad expressions while
victim and uninvolved group showed no significant differences. The results emphasized
further investigation by associating with findings with theory of mind skills and cognitive
303
affective empathy in order to explain the role of this deficit in aggressive behavior of
students involved in bullying. Unlike victims and bullies, bully-victim group did not
differ than uninvolved students in recognizing fearful expression. Despite the being most
deficient in general emotion recognition ability, better recognition of fear related cues
could explain why these students are involved in bullying. This might facilitate them in
approaching the potential target (Marsh, Ambady, & Kleck, 2005). Further studies should
consider as why children filter and select some environmental cues over others. What is
attended to by the child would determine what kind of reaction he/she may opt. Finally
the results showed that irrespective of existing mood, bullying was related to less
Hypothesis
3. There would a positive relationship between age and the quality of friendship.
Method
Participants
Students (N = 672; 319 boys, 353 girls, Mage = 10.8 years, SD = 1.08, age range:
9–14 years) were recruited from different schools of Lahore. Details of the sample are
Friendship Qualities Scale Urdu version (18 item scale) that fitted well to the
present data set (see section B; study 1) was used for analyzing the results. The final
Urdu version included four subscales namely, Trust (5 items), Intimacy (5 items),
Support (3 items) and Conflict (5 items). Each having acceptable reliabilities (see section
B, study 1). However we computed the total FQS score by adding scores on Trust,
Intimacy and Support being significantly correlated with each other and found unrelated
with Conflict subscale. Although the items in Conflict subscale were reverse scored and
pattern, adding which would have decreased the reliability of the measure. Therefore
while referring to FQS in results portion we meant the sum of three subscales mentioned
Results
Table 4.1b
Bully/Victim Status
Gender
Bully Victim Bully Victim Uninvolved Total
Boys 59 69 103 88 319
Table 4.2 b
and Grade
Source SS df MS F p η2
Results indicated a univariate main effect for bully/victim groups, F (3, 648) =
2.649, p < .05, η2= .012. It was evaluated later for multiple comparisons in post hoc
analysis. There was also a main effect for gender, F (1, 648) = 19.33, p < .001, η2= .029.
Girls reported significantly higher quality of friendship (M = 57.35, SD = 7.90) than boys
friendship quality, F (2, 648) = .31, p = ns, η2= .002. Gender and grade interaction on
friendship quality was yet found significant with a marginal effect size, F (2, 648) = 3.47
308
p < .05, η2= .008. This interaction is illustrated in figure 4.3b. All other interactions
remained insignificant.
Table 4.3b
95% CI
Table 4.3b indicated that only bull-victim group significantly differed than
uninvolved students for quality of friendship. Quality of friendship for bullies, victims
and uninvolved students was not found significantly different across groups.
309
Figure 4.1 b
friendship. Friendship quality of uninvolved students seems better than other groups.
310
Figure 4.2 b
Figure 4.2 b indicates that friendship quality of girls was better than boys
311
Figure 4.3b
There is marked difference on friendship between boys and girls studying in grade
further explored the main effects and interaction of these variables on subscales of FQS
and the conflict scale. For this purpose a 4 × 2 Multivariate Analysis of Variance was
conducted.
Table 4.4b
MANOVA showing Main Effects and Interaction of Bully/Victim Groups and Gender on
Source DV SS df MS F p η2
B/V Groups Approval 56.385 3 18.795 1.555 .199 .007
Intimacy 151.481 3 50.494 3.935 .008 .017
Help 25.511 3 8.504 1.334 .262 .006
Conflict 471.199 3 157.066 6.039 .001 .027
Gender Approval 142.904 1 142.904 11.820 .001 .017
Intimacy 472.983 1 472.983 36.863 .001 .053
Help 52.521 1 52.521 8.239 .004 .012
Conflict 55.097 1 55.097 2.118 .146 .003
B/V × Gender Approval 27.760 3 9.253 .765 .514 .003
Intimacy 30.813 3 10.271 .801 .494 .004
Help 6.083 3 2.028 .318 .812 .001
Conflict 434.380 3 144.793 5.567 .001 .025
Error Approval 8027.708 664 12.090
Intimacy 8519.613 664 12.831
Help 4232.963 664 6.375
Conflict 17269.701 664 26.009
Total Approval 325478.000 672
Intimacy 307326.000 672
Help 117384.000 672
Conflict 200942.000 672
313
Results indicated main effect for bully/victim groups on intimacy, F (3, 664) =
3.93, p < .01, η2= .01, and conflict (resolution), F (3, 664) = 6.03, p < .001, η2= .027. It
was later analyzed for group comparisons. Significant main effects of gender were found
on approval, F (1, 664) = 11.82, p < .001, η2= .017; intimacy. F (1, 664) = 6.86, p < .001,
η2= .053 and help, F (1, 664) = 8.22, p < .001, η2= .012. Girls reported significantly
higher quality of friendship on these subscales than boys. Gender and Bully/victim group
interaction on conflict (resolution) was found significant, F (2, 664) = 5.56 p < .001, η2=
.25.
Scheffe’s post hoc analysis revealed that intimacy in the friendship was
students.
314
Figure 4.4b
Groups
Girls who were not involved in bullying demonstrated better conflict resolution
than boys. When involved in bullying, girls scored lower on the scale compared with
boys.
315
Discussion
In light of above mentioned results two of our hypotheses were supported. Girls
significantly scored higher on FQS than boys. Bully-victim groups had significantly
poorer quality of friendship as compared with other three groups who did not differ in
Bully-victim group was at greater risk of having poor friendship quality explained
of victims and bullies was not different than uninvolved group. It has been recognized
that bully-victim groups could suffer from severe psychosocial aftermaths of bullying.
They were regarded as clumsy, immature and irritable (Olweus, 2006). Lack of good
intentions and lack of confidence in the relationship and thus being unable to resolve the
conflicts. These mechanisms increase the risk of aggressive behavior and decrease the
chances of being defended by peers. Aggressive acts further affect the relationship and
thus we can say that poor quality of friendship could be a predictor of victimization and
an outcome of being a bully as well. Over all the results were found somewhat
contradictory of the findings (Shin, 2012) who found that bully-victims did not differ
than uninvolved students in their friendship quality. We suggest that not all the aspects
findings.
316
gender, most of the findings are in line with the existent literature as girls generally
tended to show better friendship quality than boys (Gresham & Elliot, 1990). Furman
(1996) found that girls had the benefit of more cooperation by their close friend(s) as
compared with boys. High rates of bullying and victimization in boys could be explained
group especially. Good quality of close friendship may be the doorway for an unpopular
student to the social world of accepted and well-adjusted peers. This may reduce the
bullies, victims and uninvolved students) are also valuable as Brendgen, Vitaro, and
Bukowski, (2000) found that bullies might have same kind of friends in their circle who
possess anti-social tendencies and thus encourage the aggressive behavior. Victims on the
other hand could have friendship that were equally likely to be bullied and thus were
unable to offer the protective buffer, though intimacy and sharing still existed in the
relationship.
317
monthly income and family size was evaluated in this section by analyzing their
relationship with emotion recognition accuracy scores and quality of friendship. Lastly
the data were analyzed for possible relationships among study variables. Participants,
who were identified into different bullying roles, performed on FEEL test and had also
been assessed for qualities of friendship were evaluated in this section for relationship in
Objectives
2. The association between emotion recognition ability and friendship quality of bullies,
Results
Table 4.1c
Recognition
Variables N FQ N ER
Quality of friendship (r = .09, p < .05) and emotion recognition (r = .32, p < .001)
strength of relationship was stronger between emotion recognition and exam marks.
Better friendship quality was significantly associated (r = .13, p < .05) with school
attendance. Significant positive relationship (r = .23, p < .001) was found between
monthly income and emotion recognition ability (FEEL scores). There was also
significant yet inverse relationship between number of siblings and emotion recognition
Considering the previous results (study 3, section A & B), we examined the
relationship between friendship quality and emotion recognition that was found weak yet
significant (r = .15, p < .05). This relationship was further evaluated among bully/victim
groups. We then analyzed which of these two variables was more potential predictor of
bullying and victimization in the current sample. Data on both of the variables was available
Table 4.2c
Groups
Groups N r p
Although FQS and FEEL total scores were found to be significantly correlated (r
= .15, p < .05), most of this relationship was explained by correlation between friendship
quality and emotion recognition of the bully-victim group (r = .21, p < .05).
320
Table 4.3c
Regression Analysis for Emotion Recognition Ability and Friendship Quality on Global
Victimization (N = 370)
Model B SE β t p
Table 4.3c showed that FEEL score significant predicted experience of victimization
(global item 4 of OBVQ) accounting for 3% of variance [R2 = .030. F (367) = 6.616, p < .01].
The direction was negative indicating that poorer emotion recognition ability was associated
with higher levels of victimization. Friendship quality (FQS score) however could not prove
Table 4.4c
Regression Analysis for Emotion Recognition Ability and Friendship Quality on Global
Bullying (N = 370)
Model B SE β t p
involvement causing 3.2% of variance (R2 = .032. F (367) = 7.183, p < .001). It showed
that good friendship quality was linked to fewer chances of involvement in bullying
others. FEEL score also significantly predicted involvement in bullying behavior and the
direction was negative suggesting that poor emotion recognition ability accounted for
increase in bullying
322
Discussion
The final section of study 3 bore valuable findings with reference to Socio-
in friendship quality and emotion recognition ability was significantly associated with
Attendance was also significantly associated with friendship quality indicated that those
who had more positive friendship relation were less likely to avoid the school. Social and
emotional competencies had been previously found associated with academic success by
associated with each other. In literature, bully-victim has been found to be controversial
group. They had been reported as more irritable rejected while exhibited high self-
0esteem and confidence like bullies (Seixas, Coeelho & Nicolas-Fischer, 2013). They
actually possess a combination of characteristics that make them more prone towards
compared to friendship quality which did not predict victimization. Bullying was
analyze the social- ecological theory of emotional perception, and experiences that account for
bullying process.
processing and friendship quality. The importance of detecting the ways in which
students involved in bullying interpret social scenarios can provide a basis for
intervention programs which help the students to develop quality peer interactions in
in friendship relation, to think about more reasonable response options, to make use of
competent way.
324
Chapter V
GENERAL DISCUSSION
students. The general aim of this dissertation was to investigate the effect of mood on
emotion recognition among bully/victim groups, taking the arguments offered by social
involvement in bullying with reference to students’ average exam marks and school
(emotion recognition ability and friendship quality) could relate to the academic
performance of students. In addition to identifying gender and grade differences for the
main study variables, other demographic variables (e.g. monthly income, family size and
Results from the four studies revealed valuable information and emancipated
novel avenues for future research. In this last chapter, studies are summarized and
important results are discussed. Limitations are also presented and suggestions are
outlined to handle them in future studies. The implications of the findings are also for
theory, for interventions in bullying and for further research are suggested.
Olweus Bully Victim Questionnaire (OBVQ) and Friendship Qualities Scale (FQS). Both
325
of the versions were empirically evaluated for their content validity using three different
samples for each measure. Results from the bilinguals’ samples (for OBVQ and FQS)
demonstrated highly significant and strong correlations between Urdu and English
versions of both the scales. We then proceeded with establishing the initial psychometric
properties for OBVQ and FQS using two distinct samples of 122 and 151 students
respectively.
Results of exploratory factor analysis for OBVQ were in line with the original
structure and the existing literature on adapted versions of the scale (Schwartz et al.,
2001; Solberg & Olweus, 2003). Items assessing bullying and victimization distinctly
loaded on two factors and both the subscales showed good reliability (Bullying: α = .80;
Victimization: α = .94). Consistent with the extant literature, (Cook et al., 2010), bullying
strongly correlated with aggressive behavior subscale of CBCL (r = .78, p < .001), which
personifies the externalizing behavior. Though it was also found related with
than aggressive behavior dimension (r = .41, p < .001). Studies have also reported that
bullies used to exhibit some internalizing behavior symptoms. The reason is that some
bullies also experience victimization and are named as reactive bullies or bully-victims.
They tend to show both kind of symptoms and are at higher risk of psychopathology
(Glew, Fan, Katon, & Rivara, 2008). Being bullied or victimization was strongly
associated with anxious/depressed dimension (r = .49, p < .001), and showed non-
significant relationship with aggressive behavior (r = .15, p = ns). Findings lend support
to the convergent and divergent validity of OBVQ Urdu version. Finally on a sample of
326
817 students, OBVQ yielded adequate fit indices (CFI = .94; TLI = .92; RMSEA = .05)
Despite limited agreement found between self-reports and teacher nominations (as
cited in Smith, 2014), present study found high concordance between OBVQ and teacher
nominations (ĸ = .82, p < .001). It provided additional support to the construct and
concurrent validity, yet we suggest a careful interpretation of the result as we did not use
principal axis factoring using promax method of rotation. Pattern matrix suggested four
factors with almost adequate (Yong, Hua, & Mei, 2007) internal consistency (α >
.60).The picture was different from the 5 factor model of friendship quality offered by the
authors of FQS (Bukowski et al;, 1994). Smaller sample size (N = 151) could explain the
variations. Later, original 5 factor model and the four factor model emerged from EFA in
students. On the basis of optimal fit indices (RMR = .06, GFI = .96, CFI = .95, TLI = .94,
RMSEA = .03) and higher reliability estimates (all α values > .60), we retained the CFA
multiplier (Toe et al., 2013). FQS Urdu version presented a distinctive pattern of four
subscales (Approval, Intimacy, Help and Conflict) compared with existing evidence from
original and translated versions (Fonzi et al., 1997). Younger age group (Mage= 10.86;
SDage = 1.08) of the current sample and cultural context may account for the diversity in
results.
327
Findings of study 2 (Chapter II) imply the details acquired from the full Olweus
bullying and victimization and its different types across gender and three grades (4th, 5th
and 6th) have been delineated. With regard to bully/victim roles only 34.6 % remained
uninvolved and majority of others (28.8%) were found as both perpetrators and the
girls while previous findings reported an opposite trend (Seals & Young, 2003). Students
in lower grade (4th) reported being victimized more often than others (5th or 6th graders).
Results comply with the available reports on grade differences (Salmivalli, 2002).Verbal
and relational (exclusion and rumor spreading) were the most frequently occurring forms
2004). Gender differences in types of bullying were found comparable to the existing
research (Card et al., 2008) as relational bullying (social exclusion/rumor spreading) was
more commonly reported by girls as compared with boys who reported being involved in
more direct (verbal, physical, damage to property) forms. Among other important results,
(2 to 3 students), usually lasted for a couple of weeks, and was most often reported to
school staff. Girls showed more pro-victim attitude compared with boys. Findings should
studies (Glew et al., 2008) we found that students who were part of the bullying process
While searching for the resources with reference to mood induction procedure
(MIP) for children; no indigenous protocol was found and available techniques employed
stimuli that could be culturally incompatible. This led us to develop indigenous MIP
using suitable stimuli (study 4; section A: phase I). It was essentially required before
conducting the main study on effect of mood on emotion recognition of bullies and
victims. Four techniques (Video song, Instrumental music, Emotive pictures, and
for each technique. These techniques were then empirically tested on a student sample
ANOVA found that the all the techniques produced significant changes in existing mood
in an enviable way. Pre and post manipulation results revealed that video song was the
most efficient technique in both happy and sad condition followed by a combination of
pictures presented with background music as compared with music and pictures alone.
Findings are consistent with earlier studies suggesting significant mood change as a result
of music and emotive pictures (Goritz & Moser, 2006; Pagliaccio et al., 2012). In the
(N = 394) from the main sample of study 2. Mood priming (using video song along with
picture/music combination) proved successful for sad, happy and neutral conditions. The
study failed to support the mood congruency hypothesis in general. Yet performance of
detecting the angry expressions. Observing the error profiles of victimized students could
explain the information processing in sad and happy moods. Although all the students
students, deficits were evident for bully-victim and victim groups. Results compliment
the deficit model of SIP with regard to bullying (Crick & Dodge, 1994). Prospective
processing to see if there is a stable pattern of deficits for children who are victimized.
(Espeealage & Swearer, 2003), results of study 4 (Chapter 3; section B) revealed that
significantly better friendship relation than bullies, victims and especially bully-victims
who scored lowest on FQS. Major differences emerged for conflict resolution and
intimacy subscales. Over all girls demonstrated better quality of friendship than boys.
There is need to develop quality peer support networks not only to prevent bullying but
friendship quality as important precursors to bullying and victimization, this section (C)
Academic cost of being involved in bullying is evident (Juvonen et al., 2011) and
supported in study 2. Poor emotion recognition skills also predicted educational outcome
school avoidance (low attendance percentages). The results are grounded on extant
330
literature (Altermatt & Pomerantz, 2005; Goetz et al., 2005). Monthly income was
significantly correlated with emotion recognition yet not with friendship quality.
Additionally, friendship quality and emotion recognition ability were correlated though
not strongly (r = .15, p < .05); might explain the mechanisms that link the individual
capacities and processes with the ways in which the social experiences become embedded
in the individual.
Although current research has not further addressed these important links, yet we
suggest that children possess complex cognitive and social abilities that, particularly
when applied to affective information received from the environment, may provide
insight into both behavioral adaptation and mal-adaptation. Finally, the present work is an
important area to advance, as shown by recent attempts to clarify the role of emotion in
existing theoretical frameworks and the growing recognition of their importance in social
The studies presented in the thesis contribute to our knowledge in the field of
bullying at school and fill certain gaps identified in our review of literature. Here we
place of our findings within the theoretical framework and prospect research.
resulted in functional characteristics of the phenomenon. Students were able to take into
concern the basic elements of power imbalance, repetition, harmful intention and various
embryonic stage; therefore a deliberate insight offered by the first section of study 1
guides the researchers to be cautious while assessing the construct of bullying. It also
adds useful information to the growing body of the cultural context of bullying that has
been recently highlighted by researchers (Smith & Monks, 2008; Smith, 2014).The study
supports the use of definition based measure which may distinguish bullying from other
forms of aggression.
has serious mental health implications (Stagg & Sheridan, 2010). These
Adeleke, Omoike, and Afen-Akpaida (2008). Tendency for increased somatic complaints
can add to the stress and expose the students to serious illness (Aluede, 2006). A detailed
physical health status. Not only bully/victim groups, the bystanders may also feel
insecurity and discomfort as a result of witnessing violent behavior (Polanin, Espelage, &
Pigot, 2012) Schools should not take the issue for granted and adapt the measures for
Majority of our sample involved in bullying were bully-victims and more boys
reported being bullied. We later found that bully-victim was more vulnerable to socio-
affective repercussions (poor friendship quality and emotion recognition skills). Bully-
victims have been thought to be a distinct and high risk group by researchers having more
mental health and behavior issues, and reporting greater difficulty in peer relationships
(Andreou, Valochou & Didaskalou, 2005; Renda, Vassallo, & Edwards, 2011). This
distinct group could be studied in detail for further clarity of processes that account for
332
their pro-bully attitude and peer rejection aspects. Moreover being girls equally involved
in overall bullying and more prone towards relational bullying needs further reflections
on affective and cognitive empathy that might serve as moderator to this relationship
during data collection, that most of the teachers were unfamiliar to the word ‘bullying’.
Awareness itself can prevent bullying as teachers and peers would be more vigilant to the
can address the issue besides a carefully planned anti-bullying strategy. Studies have also
recommended collaborative efforts between parents and teachers in order to help the
victims and snub bullying (Ahmed & Braithwaite, 2004). It is important for teachers to
communicate and note the awareness of the parents in bullying behaviors or victimization
of their children. Moreover, children who act as uninvolved could also be trained to
become defenders. This research can be an initiative to nip the evil in the bud by
identifying the potential source of violence at initial level, which later could be harmful
margin that, when accurately processed, can lead to an empathic response. In this regard,
both internal and external dispositions may persuade the understanding of contextual
cues. Overall findings of the present research (in study 3) did not support the existing
evidence for effects of positive or negative mood on social perceptions and judgments
et al. (2007). Thus neither mood congruence nor incongruence was found significant
while differences in emotion recognition skills were not regulated by mood states.
diverse findings; however the results afford implications for cognition-affect interface
which discusses the role of mood for processing evidence related to interpretation of
emotional cues in a social situation. Furthermore, basic model of mood congruency for
cognition and emotion interplay could be insufficient in capturing the full richness of the
issue. Variations can be explained by more deliberate studies including eye trackers,
fMRIs or other imaging techniques. In addition, deficient groups may entail more
emotions.
of students. Knowing the nature of emotion recognition deficits in bully/victim group can
guide be a helpful in designing such programs. Emotion recognition skill can be learnt
(Silver & Oakes, 2001) and smart phone applications (Lamm et al., 2014) that have been
used to teach emotion recognition skills to autistic children can also be used with
bully/victims who suffer from the deficit. Such innovative technology can also be used to
Along with cognitive and emotional issues, there are also social problems related to
bullying and victimization. Based on our finding from study 1, we suggest that patterns of
334
different from the original framework of friendship quality. Moreover, findings of study 4
Childhood peer relations have certain implications for optimistic outcomes in for social
competence in later life. Implementing peer networks that assist peer interactions and
(Cowie & Olafsson, 2000).Taking the relationship between friendship quality and
emotion recognition into account, emotional knowledge can help in improving positive
social behaviors. Creating dyadic links between bullies and victims by means of work
This is the first known study to discuss a variety of aspects of bullying behavior
gender and grade differences, cognitive-affective and social facets and the academic
corollaries. While the research filled some gapes, it left some anomalies as well. The
topic can be expanded to other steps of SIP and reciprocal peer interaction with special
Ethical Considerations
participants and users. With regards to the ethical considerations, the present research
1. Permission was sought from the authors of OBVQ and FQS to use and
translate the measures into Urdu. Permission was also obtained from the
6.3.
their schools. Objectives were explained to them and the researcher took
the responsibility for not violating the rules during data collection. Some
schools discontinued after few sessions and did not share their academic
records or attendance. They were not forced in this regard and their rights
4. All the participants were informed about the aims and objectives of the
studies. Consent was sought from parents, however students were not
any stage. They were free from deception or stress that might arise from
6. The respondents were also assured protection through privacy and all
Names were replaced with codes while entering the data to SPSS.
336
watching some materials and if they found it uncomfortable they could tell
the experimenter about this and might discontinue at any time. Moreover
through happy MIP. Mood repair after emotion recognition task had been
completed.
8. Counseling services were offered by the researcher and parents were also
provided to the school authorities and the parents to consult further about
any participant. Some of the schools celebrated anti bullying week at their
schools.
There is no study without limitations and the present research also encountered a
few. Most of these limitations came from the sources of time and consent. Here are some
1. The data were not obtained from the whole class(es) as 80% of the participant
schools did not allow for this. Although the students were randomly selected from
the classes, whole class administration might have resulted different prevalence
estimates. Gathering a few students separately in a room might have biased their
However, teacher nominations were also obtained for these students that showed
high concordance with self-reports. Researchers are suggested to use whole class
2. Another limitation regarding OBVQ in the present research was not following the
many of the students still mentioned their names on the questionnaire, which
reflected that they were not concerned of being exposed to others or reluctant any
way to report their identification. Another reason was that participants of study 4
were recruited from the sample of study 3. Therefore and coding procedure was
avoided that might had created confusions resulting in mismatch of data sets.
Nevertheless the students were assured of the confidentiality of the data and the
names were replaced with codes while entering the data to SPSS.
3. Later research can define the uninvolved group more explicitly in terms of their
4. Only two subscales of Child Behavior Checklist (CBCL; Achenbach, 1991) were
investigated initial psychometric properties of the measure. CBCL full scale could
and externalizing behavior problems. Similarly Social competence full scale and
other measures of peer relationship quality could have yielded clearer picture of
5. Factor structure of FQS represented a distinct pattern for FQS Urdu version. This
variation could be attributed to smaller sample size as larger samples yield better
done using FQS. The questionnaires were administered to the whole class and
only a few randomly selected students completed the scales, therefore friendship
7. Experiments were not conducted in laboratory as it was not possible to bring all
the children to university. Parents also did not allow for this. Security issues and
8. We employed two MIPs which is not unusual for mood induction researches, and
that too proved to be very effective in terms of altering mood states. Yet this
might influenced mood congruence. Future researchers can use alternate methods
such as film clips. Additionally, pictures and music excerpts should have been
9. Data on FEEL were analyzed only for accuracy scores and reaction time. Separate
profiles of error patterns for each of the bully/victim groups could better elucidate
10. FEEL test only measures recognition for static and basic emotions. Presenting
scenarios from original life including dynamic and more complex emotions such
as shame, guilt, etc could bear more fruitful results and implicate for intervention
processing.
data from almost 200 participants, yet it proved very time consuming and
that only few schools agreed to participate in study 4 and they too offered a tight
schedule that had to be met before annual exams. Keeping in view the time
constraints we stopped assessing IQ and had to exclude this variable from main
analysis.
12. General mood could serve as a moderator in determining the effect of induced
friendship quality, the issue is left unexplained in the current research and needs
14. Data on certain demographic variables such as family income, parent’s education
and occupation, family size etc was missing as participants did not provided
share academic records (marks and attendance) resulting in limited data on these
variables.
340
15. This study was cross-sectional. Longitudinal studies are required to explain
Conclusions
scant. Present research concurrently addressed these dynamics in four studies. Each study
The research provided validated Urdu versions of two widely used instruments
(OBVQ and FQS) for measuring maladaptive and positive aspects of peer relationships:
protocol using familiar and age appropriate stimuli is another addition to the rarely
emotions and the existing mood hampered the emotion recognition accuracy of
victimized students. Findings further indicated that poor quality of friendship could
research predominantly emphasize the need for emotional and social skill training for
phenomenon could enhance the likelihood of anti-bullying and pro-social attitude among
level are contemplative for all the stakeholders. For students, it is imperative to realize
their role and possible effects of bullying on themselves and others. School authorities
should create healthier school climate and better academic outcomes. Parents must
acknowledge the issue and address it timely in order to prevent psychosocial aftermaths.
determinants of bullying; and most importantly for government agencies must establish
REFERENCES
Achenbach, T. M. (1991). Manual for Child Behavior Checklist/ 4-18 and 1991 Profile.
Adair, V. A. Dixon, R. S., Moore, D. W., & Sutherland, C. M. (2000). Ask your mother not
Adnan, A., Chaudhry, A. A., &Malik, M. I. (2012). Emotional Intelligence and Students
Ahmed, M., Hussain, I., Ahmed, M., Ahmed, S., & Tabassum, R. (2012). Impact of Bullying
3, 13-22.
Ahmed, E., & Braithwaite, V. (2004). Bullying and victimization: Cause for concern for both
Ahmer, S., Yousafzai, A. W., Bhutto, N., Alam, S., Sarangzai, A. K., & Iqbal A. (2008).
Alikasifoglu, M., Erginoz, E., Ercan, O., Uysal, O., & Albayrak-Kaymak, D. (2007).
among high school students in Istanbul, Turkey. European Journal of Pediatrics, 166,
1253-1260.
Alikasifoglu, M., Erginoz, E., Ercan, O., Uysal, O., Albayrak-Kaymak, D., &Iiter, O. (2004).
343
Violent behaviour among Turkish high school students and correlates of physical
81.
Aluedo, O., Adeleke, F., Omoike D., & Afen-Akpaida, J. (2008) A review of the extent,
Andershed, H., Kerr, M., & Stattin, H. (2001). Bullying in school and violence on the streets:
Are the same people involved. Journal of Scandinavian Studies in Criminology and
Andreou, E., Vlachou, A., & Didaskalou, E. (2005). The roles of self-efficacy, peer
Anjum, N., & Malik, F. (2010). Parinting Practices in Mother of Children with ADHD: Role
Atik, E., Cok, F., Coban, F., Dogan, A., & Karaman, G. N. (2014). The Turkish Adaptation
Backs, R.W., Silva, S.P., & Han, K. (2005). A comparison of younger and older adults’ self-
Baldry, A. (2004). The impact of direct and indirect bullying on the mental and physical
Ball, H. A., Arseneault, A., Taylor, A., Maughan, B., Caspi, A., &Moffitt, T. E. (2008).
childhood. Journal of Child Psychology and Psychiatry and Allied Disciplines, 49,
104-112.
Barboza, G. E., Schiamberg, L. B., Oehmke, J., Korzeniewski, S. J., Post, L. A., & Heraux,
10.1177/0013164403251333.
Barden, R. C., Garber, J., Duncan, S. W., &Masters, J. C. (1981). Cumulative effects of
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, & IJzendoorn, M. H. (2007).
Barnett, M. A., King, L. M., & Howard, J. A. (1979). Inducing affect about self or other:
Batsche, G. M., & Knoff, H. M. (1994). Bullies And Their Victims: Understanding a
Pervasive Problem in the Schools. School Psychology Review, 23, 165 – 175.
doi:10.1080/15388220801973912.
Baumgartner, T., Lutz, K., Schmidt, C. & Jäncke, L. (2006). The emotional power of music:
How music enhances the feeling of affective pictures. Brain Research, 1075(1),151-
164.
Beaton, D. E., Bombardier, C., Guillemin, F., & Ferraz, M. B. (2000). Guidelines for the
3191.
http://www. cyberbullying.us/cyberbullying_girls_victimization.pdf.
Benbenishty, R., & Astor, R. A., (2005). School violence in context: Culture, neighborhood,
107(2), 238-46.
Bentler, P. M., & Dijkstra, T. K. (1985). Efficient estimation via linearization in structural
North-Holland.
Bentley, K.M. & Li, A. K. F. (1995). Bully and victim problems in elementary schools and
students' beliefs about aggression. Canadian Journal of School Psychology, 11, 153-
165.
Beran, T. N., & Lupart, J. (2009).The Relationship between School Achievement and Peer
doi:10.1177/014303430810185.
Beran, G., & Stewat, S. (2008). Children’s experience of loneliness at school and its relation
to bullying and the quality of teacher interventions. The Qualitative Report, 9, 483-
499.
Berndt, T. J. (1998). The features and effects of friendship in early adolescence. Child
Berthold, K., & Hover, J. (2000). Correlates of bullying and victimization among
21(1), 65-78.
Besag, V. (1989). B ullies and Victims in Schools by Valerie Besag. 232, McGraw-Hill.
Bilello, M., Casiglia, A. C., Lo Coco, A., & Miceli, S. (1995). Il Ruolo Della Competenza
Boivin, M., Hymel, S., & Bukowski, W. M. (1995). The roles of social withdrawal, peer
Bollen, K., & Long, J. (1993). Testing Structural Equation Models. Sage Publications,
Bollmer, J. M., Milich, R., Harris, M. J., & Maras, M. A. (2005). A friend in need: the role of
Bouhyys, A. L., Bloem, G. M., & Groothuis, T. G. (1995). Induction of depressed and elated
Boulton, M. J. (1997). Teachers’ views on bullying: Definitions, attitudes and ability to cope.
835X.1994.tb00637.x.
Boulton, M. J., & Underwood, A. (1992). Bully/victim problems among middle school
Bowen, G. L., & Holcom, J. M. (2000). Social support networks and school outcomes: The
centrality of the teacher. Child & Adolescent Social Work Journal, 17(3), 205-226.
Bowes, L., Arseneault, L., Maughan, B., Taylor, A., Caspi, A.,& Moffitt, T. E. (2009).
School, neighborhood and family factors are associated with children’s bullying
349
Bracken, B. A., & Barona, A. (1991). State of the art procedures for translating, validating
Bradley, B. P., & Mogg, K., & Lee, S. C. (1997). Attentional biases for negative information
35(10), 911-927.
Bradley, M. M., & Lang, P. J. (1994). Measuring Emotion: The Self-Assessment Manikinand
25(1), 49-59.
Brendgren, M., Barker, E. D., Boivin, M., Fontaine, N., Arseneault, L., Vitaro, F.,
Bissonnette, C., & Trembley, R. E. (2011). Predictive Validity and Early Predictors
Brendgen, M., Boivin, M., Vitaro, F., Girard, A., Dionne, G., & Pérusse, D. (2008). Gene-
Brendgen, M., Vitaro, F., & Bukowski, W. M. (2000). Stability and variability of adolescents
Development, 9, 205–225.
Brendgen, M., Vitaro, F., & Bukowski, W. M. (2000). Deviant friends and earlyadolescents’
173–189.
350
2680.4.3.264.
style and perceived tempo. Journal of Research in Music Education, 48, 237-248.
Brunstein-Klomek, A., Marrocco, F., Kleinman, M., Schonfeld, I., & Gould, M. (2007).
Brown, D. W., Riley, L., Butchart, A., & Kann, L. (2008). Bullying among youth from eight
African countries and associations with adverse health behaviors. Pediatric Health,
2(3), 289-299.
Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In: K. A. B
ollen & J. S. Long (Eds.), Testing structural equation models (pp. 136-162).
Brunye, T. T., Mahoney, C. R., Augustyn, J. S., & Taylor, H. A. (2009). Emotional state and
Bukowski, W. M., Hoza, B.. & Boivin, M. (1996). The company they keep: Friendship in
friendship among early adolescent boys and girls. The Journal of Early Adolescence,
Bukowski, W.M., Hoza, B., & Boivin, M. (1994). Measuring friendship quality during pre-
Friendship Qualities Scale. Journal of Social and Personal Relationships, 11, 471-
484.
Bukowski, W. M., Hoza, B., & Newcomb, A. F. (1991). Friendship, popularity, and the 'self'
Department of Psychology.
Burks, V. S., Laird, R. D., Dodge, K. A., Pettit, G. S., & Bates, J. E. (1999). Knowledge
Development, 8, 220–236.
Burk, F.L. (1897). Teasing and bullying. Pedagogical Seminary, 4, 336 -371
Cairns, R. B., & Cairns, B. D. (1994). Lifelines and risks: Pathways of youth in our time.
Camodeca, M., Goossens, F. A., Schuengel, C., & Terwogt, M. M. (2003). Links between
Camodeca, M., Goossens, F. A., Terwogt, M., & Schuengel, C. (2002). Bullying and
victimization among school-age children: Stability and links to proactive and reactive
Caravita, S. C. S., Di Blasio, P., & Salmivalli, C. (2010). Early adolescents' participation in
352
10.1177/0272431609342983.
Caravita, S. C. S., DiBlasio, P., & Salmivalli, C. (2009).Unique and interactive effects of
empathy and social status on inovlvement in bullying. Social Development, 18, 140–
163.
Card, N. A., Stucky, B. D., Sawalani, G. M., & Little, T. D. (2008). Direct and indirect
79(5), 1185–1229.
Dominated Great Lakes Coastal Wetland. Natural Areas Journal, 29(3), 263-280.
doi: http://dx.doi.org/10.3375/043.029.0306.
Carlyle, K. E., & Steinman, K. J. (2007). Demographic differences in the prevalence, co-
77, 623-629.
Carrera-Bastos, P., Fontes-Villalba, M., James, H. O., Keefe, E., Lindeberg, S., & Cordain,
L. (2011). The western diet and lifestyle and diseases of civilization. Research
doi: http://dx.doi.org/10.2147/RRCC.S16919.
Center for Disease Control and Prevention (2014). Suicide prevention: youth suicide.
Centers for Disease Control and Prevention. (2012). Youth Risk Behavior Surveillance
www.cdc.gov/mmwr/pdf/ss/ss6104.pdf.
Centers for Disease Control and Prevention. (2011). Youth risk behavior surveillance—
353
United States, MMWR Surveillance Summaries 2012; 61(no. SS-4). Reterived from
www.cdc.gov/mmwr/pdf/ss/ss6104.pdf.
Center for Disease Control (2010). Leading cause of death reports 1999-2007 [data file].
Chamberlain, T., George, N., Golden, S., Walker, F., & Benton, T. (2010). Tellus 4 National
http://publications.education.gov.uk/eOrderingDownload/DCSF-RR218.pdf.
Charach, A. Pepler, D., & Ziegler, S. (1995). Bullying at School. Education Canada, 5. 81-
95.
Chavez, L. M., & Canino, G. (2005). Toolkit on translating and adapting instruments.
Cheng, M., & Ward, J. (2004). The Ottawa-Georgia Mood Scales. Retrieved from
http://www.docstoc.com/docs/88073630/Ottawa-Georgia-Mood-Scales.
Chepenik, L. G., Cornew, L. A., & Farah, M. J. (2007). The Influence of Sad Mood on
3542.7.4.802.
11.65.
Chow, C., Ruhl, H., & Buhrmester, D. (2013). The mediating role of interpersonal
Coloroso, B. (2004). The bully, the bullied, and the bystander. New York, NY: Harper
Ciarrochi, J., & Heaven, P. C. L. (2012). Religious Values and the Development of Trait
354
Hope and Self-Esteem in Adolescents. Journal for the Scientific Study of Religion,
Ciucci, E., Smorti, A., Fonzi, A. (1997). II rapporto bullo-vittima in soggetti di scuola media:
Cohen, G. L, Garcia, J., Apfel, N., &Master, A. (2006). Reducing the Racial Achievement
Cook, C. R., Williams, K. R., Guerra, N. G., & Kim, T. E. & Sadek, S. (2010). Predictors of
Conklin, C. A., & Perkins, K. A. (2005). Subjective and reinforcing effects of smoking
Conrad, K. J., Wright, B. D., McKnight, P., McFall, M., Fontana, A., & Rosenheck, R.
(2004). Comparing traditional and Rasch analyses of the Mississippi PTSD Scale:
15-30.
Cornell, C. M., Sheras, P., & Cole, M. (2006). Identification of School Bullies by Survey
Costello, A. B. & Osborne, J. (2005). Best practices in exploratory factor analysis: four
recommendations for getting the most from your analysis. Practical Assessment
Coulton, C. J., Korbin, J. E., Su, M., & Chow, J. (1995). Community level factors and child
Cowie, H. (2000). Bystanding or standing by: Gender issues in coping with bullying in
Craig, E. W., Jasmine, E. S., Warwick, F., Vincent, C., John, P., Smold, L., & Oceanogr.
Craig, W. M., Pepler, D., & Atlas, R. (2000). Observations of bullying in the playground and
Craig, E. W., Peters, R. and Konarski, R. (1998). Bullying and Victimization Among anadian
School Children. TheWorking Paper Series for the Applied Research Branch, Human
Craig, W. & Pepler, D. (1997). Observations of bullying and victimization in the schoolyard.
Crick, N.R., & Dodge, K.A. (1999). Superiority in the eye of the beholder: A comment on
Crick, N. R., Werne, A., Casas, J. F., & Ku, H. C. (1999). Relational and physical forms of
Crick, N. R., & Bigbee, M. A. (1998). Relational and overt forms of peer victimization: A
347.
Crick, N. R., & Grotpeter, J. K. (1995). Relational aggression, gender, and social-
Crick, N. R., & Dodge, J. K. (1994). Relational aggression, gender, and social–psychological
356
doi: 10.1002/j.1556-6678.2004.tb00338.x.
Cross, D., Shaw, T., Hearn, L., Epstein, M., Monks, H., & Lester, L., (2009). Australian
www.deewr.gov.au/Schooling/NationalSafeSchools/Pages/research.aspx.
Cummings, A. J., & Rennels, J. L. (2013). How Mood and Task Complexity Affect
doi: 10.1111/sode.12038.
Currie, C., Nic-Gabhainn, S., Godeau, E., Roberts, C., Smith, R., & Currie, D. (2008).
Inequalities in young people's health: HBSC international report from the 2005/06
Survey. Health Policy for Children and Adolescents, No. 5, WHO Regional Office for
Dake, J., Price, J., & Telljohann, S. (2003). The na- ture and extent of bullying at school. The
Darwin, C. (1965). The expression of the emotions in man and animals. University of
Deldin, P. J., Deveney, C. M., Kim, A. S., Casas, B. R., & Best, J. L. (2001). A slow wave
Evidence from Head Start. American Economic Journal: Applied Economics, 1(3),
111–134.
Demir, M., & Urberg, K. A. (2004). Friendship and adjustment among adolescents.
Demmings, J. L. (2009). Examining Ethnic Identity and Friendship Quality among High
Demorest, S. M., & Schultz S. J. M. (2004). Children's Preference for Authentic versus
Denham, S. A., Von Salisch, M., Olthof, T., Kochanoff, A., & Caverly, S. (2002).
Emotions and peer relationships. In C. Hart & P. K. Smith (Eds.), Handbook of child
Denham, S. (1998). Emotional development in young children. New York: Guilford Press.
Department of Health and Human Services. (2009). Obesity in children: New prevention
DeVoe, J., & Kaffenberger, S. (2005). Student Reports of Bullying: Results from the 2001
DC.
Dijkstra, P. U., Wilgen, C. P., Versteegen, G. L., & Fleuren, M. J. T. (2009). Chronic pain
358
Doğan, T., Karaman, N. G., Coban, W. (2012). A Long-term study of the counseling needs of
Dodge, J. P. (1986). Mood and judgment: The Affect Infusion Model (AIM). Psychological
Dolgin, K. G., & Adelson, E. H. (1990). Age changes in the ability to interpret affect in sung
Due, P., Damsgaard, M.T., Lund, R., & Holstein, B. E. (2009). Is bullying equally harmful
for rich and poor children. European Journal of Public Health, 19(5), 464-469.
Due, P., Holstein, E. B., Lynch, J., Diderichsen, F., Gabhain, S. N., Scheidt, P., &Currie, C.
Duck, L. (2005). Publishing models and article dates explained. International Journal of
Duncan., R. D. (1999). Peer and Sibling Aggression: An Investigation of Intra- and Extra-
doi: 10.1177/088626099014008005J.
Dunn, J. A. (2004). Anxiety, stress, and the performance of complex intellectual task: A new
look at an old question. Journal of Consulting and Clinical Psychology, 32, 669-673.
359
Dunn, J. A., & Cutting, A. L. (1999). Understanding Others, and Individual Differences in
doi: 10.1111/1467-9507.00091.
Dupper, D. R.. & Meyer-Adams, N. (2002). Low-level violence: A neglected aspect of school
Dussich, J. P. J., & Maekoya, C. (2007). Physical child harm and bullying-related behaviors:
A comparative study in Japan, South Africa and the United States. International
Dwyer, P., Forsyth, B., King, L., & Seetaram, N. (2010). Migration Related Determinants of
Eaton, D. K., Kann, L., Kinchen, S., Shanklin, S., Flint, K. H., Hawkins, J., Wechsler, H.
(2012). Youth risk behavior surveillance: United States, 2011. Morbidity and
Eisenberg, N., & Fabes, R. A. (1992). Emotion, regulation, and the development of social
Ekman, P., & Davidson, R. J. (1994). The nature of emotion: Fundamental questions. New
Ekman, A. (1973). Cross culture study of facial expression, In P. Ekmon (Ed), Darwin and
Erath, S. A., Flanagan, K. S., & Bierman, K. L. (2008). Early adolescent school adjustment:
Associations with friendship and peer victimization. Social Development, 17, 853-
870.
Erber, R., & Erber, M. W. (1994). Mood and processing: A view from a self-regulation
Eslea, M., & Smith, P. K. (2000). Pupil and parent attitudes towards bullying in primary
Espelage, D. L., Low, S., Rao, M. A., & Hong, J. S., (2013). Family violence, bullying,
and sexual orientation among high school students: What influence do parents and
Espelage, D. L., & Holt. (2007). Willingness to intervene in bullying episodes among middle
32, 776–801.
205-220.doi: 10.1111/1467-8624.00531.
Espelage, D. L. & Swearer, S.M. (2003). Bullying prevention and intervention. Full-day
Espelage, D. L. (2002). Bullying in early adolescence: The role of the peer group. ERIC
361
Espelage, D. & Holt, M. (2001). Bullying and victimization during early adolescence: Peer
Espelage, D. L., & Asidao, E. (2001). New Perspectives on Bullying Prevention: Why are
Espelage, D. L., Bosworth, K., & Simon, T. R. (2000). Examining the social context of
326-333.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating the
4(3), 272-299.
Fantuzzo, J., Sekino, Y., & Cohen, H.L. (2004). An examination of the contributions of
interactive peer play to salient classroom competencies for urban head start children.
Fantuzzo, J., Perry, M. A., & McDermott, P. (2004). Preschool approaches to learning and
Fairclough, S. H. Van der Zwaag, M., Spiridon, E., & Westerink, J. (2014). Effects of mood
Farrington, D., & Baldry, A. (2010). Individual risk factors for school bullying. Journal of
Farrington, D. P. (1993). Understanding and preventing bullying Crime and Justics, 17, 381-
458.
Fekkes, M., Pijpers, F. I. M., & Verloove-Vanhorick, S. P. (2004). Bullying behavior and
17–22.
Field, A. (2013). Discovering Statistics using IBM SPSS Statistics. University of Sussex.
Field, T., Diego, M., & Sanders, C. (2002). Adolescents' parent and peer relationship.
Adolescence, 37,121-122.
Fitzpatrick, K. M., Dulin, A. J.,& Piko, B. F. (2007). Not just pushing and shoving: School
bullying among African American adolescents. Journal of School Health, 77(1), 16-
22.
363
Flouri, F., & Buchanan, A. (2003). The Role of Mother Involvement and Father Involvement
doi: 10.1177/0886260503251129.
Floyd, F. J., & Widaman, K. F. (1995). Factor analysis in the development and refinement of
Fonzi, S., Schneider, B. H., Tani, F., & Tomada, G. (1997). Predicting children’s friendship
status from their dyadic interaction in structured situations of potential conflict. Child
Forero, R. (2013). Bullying behaviour and psychosocial health among school students in New
South Wales, Australia: Ross sectional survey. British Medical Journal, 319, 344-34.
Forgas, J. P. (1995). Mood and judgment: The Affect Infusion Model (AIM). Psychological
Forgas, J. P., Burnham, D. & Trimboli, C. (1988). Mood, memory and social judgments in
Fox, C. (2004). Evaluating the effectiveness of a social skills training programmer for
French, D. C., Bae, A., Pidada, S., & Lee, O. (2006). Friendships of Indonesian, S. Korean
French, D. C., Lee, O., & Pidada, S. U. (2006). Friendships of Indonesian, South Korean and
Frey, K. S., Hirschstein, M. K., Snell, J. L., Edstrom, L., Van Schoiack, L., MacKenzie, E. P.,
479–491.
364
Frost, L. (1991). A primary school approach-what can be done about the bully.In: Elliot,
Fu, Q., Land, K. C., & Lamb, V. L., (2012). School Bullying, Socioeconomic Status and
Furlong, M. J., Sharkey, J. D., Felix, E., Tanigawa, D., & Greif-Green, J. (2010). Bullying
Newcomb, & W.W. Hartup (Eds). The copany they keep: Friendship in childhood and
Gasper, K. (2003). When necessity is the mother of invention: Mood and problem solving.
Gasper, K., & Clore, G. L. (2002). Attending to the big picture: Mood and global versus local
Gallardo, P., Baños, R., Belloch, F. M., & Rupierez, A. (1999). Attentional biases and
George, T. P. (1996). Friendship networks of unpopu- lar, average, and popular children.
Gerrards-Hesse, A., Spies, K., & Hesse, F. W. (1994). Experimental inductions of emotional
Glew, G. M., Fan, M., Katon, W., & Rivara, F. P. (2008). Bullying and school safety. The
acceptance, friendships, and peer networks. Journal of School Psychology, 41, 235-
284.
Gini, G., & Pozzoli, T. (2009). Association between bullying and psychosomatic problems:
Gini, G, Albiero, P., Benelli, B., &Altoè, G. (2007). Does empathy predict adolescents'
Bullying Surveillance Among Youths: Uniform Definitions for Public Health and
Recommended Data Elements, Version 1.0. Atlanta, GA; National Center for Injury
Prevention and Control, Centers for Disease Control and Prevention and U.S.
Glew, G. M., Fan, M., Katon, W., & Rivara, F. P. (2008). Bullying and school safety. The
Glew, G. M., Fan, M., Katon, W., Rivara, F. P., & Kernic, M. A. (2005). Bullying,
1649.25.2.277.
Goetz, T., Frenzel, A. C., Pekrun, R., & Hall, N. (2005). Emotional intelligence in the context
Huber Uublishers.
Goffin, R. D. (2007). Assessing the adequacy of structural equation models: Golden rules and
10.1177/0886260511431436.
Gorman, A.H., Dodge, K.A., &Bates, J.E. (2008). Friendships with peers who are low or
high in aggression as moderators of the link between peer victimization and declines
Göritz, A.S. (2007). The induction of mood via the WWW. Motivation and Emotion, 31, 35–
47.
Göritz, A.S., & Moser, K. (2006). Web-based mood induction. Cognition and Emotion, 20,
887–896.
Gottfredson, G.D., Gottfredson, D.C., Payne, A., & Gottfredson, N.C. (2005). School
Graham, S., Hudley, C., & Williams, E. (1992). Attributional and emotional determinants of
Gresham, F. M., & Elliott, S. N. (1990). Social skills rating system. Circle Pines, Minnesota:
Halberstadt, J. B., & Niedenthal, P. M. (1997). Emotional state and the use of stimulus
1033.
Hanif, R., Nadeem, M., & Tariq,S. (2011). Bullying in Schools: Attitudes of Children,
Hanif, R. (2008). Perceptions and Attitudes towards Bullying and School Social Climate: A
(UK).
Hanson, T., Austin, G., & Zheng, H. (2010). Academic performance and school well-being in
California. California Healthy Students Research Project Research Brief. No. 1. San
Francisco,CA: WestEd.
Harper, B. D., Lemerise, E. A., & Caverly, S. L. (2010). The effect of induced mood on
Hartup, W. W., & Abecassis, S. (2002). Friends and enemies: Their significance in child
Hawkins, D. L., Pepler, D., & Craig, W. M. (2001). Peer interventions in playground
doi: 10.1111/1469-7610.00629.
368
Haynie, D. L., Nansel, T., Eitel, P., Davis Crump, A., Saylor, K., Yu, K., & Simmons-
Hazemba, A., Siziya, S., Muula, A. S., & Rudatsikira, E. (2008). Prevalence and correlates
of being bullied among in-school adolescents in Beijing: results from the 2003 Beijing
doi:10.1186/1744-859X-7-6.
Hazler, R. J., Miller, D. L., Carney, J. V., Green, S. (2001). Adultre cognition of school
Heald, T. R. (1994). Judgement in the case between R. H. Walker and Derbyshire County
522.
Halberstadt, A. G., Denham, S. A., & Dunsmore, J. (2001). Affective social competence.
Hertz, M. F., David-Ferdon, C. (2008). Electronic media and youth violence: A CDC issue
brief for educators and caregivers. Centers for Disease Control, 6, 1-22.
Hills, P. J., Werno, M. A., & Lewis, M. B. (2011). Sad people are more accurate at face
Hilton, A., & Skrutkowski, M. (2002). Translating instruments into other languages:
Hinduja, S., & Patchin, J. W. (2010). Bullying, Cyberbullying, and Suicide. Archives of
Hinduja, S., & Patchins,W. P. (2009). Cyberbullying reesearch Summary Cyberbullying and
Hodges, E. V. E., Boivin, M., Vitaro, F., & Bukowski, W. M. (1999). The power of
Holt, M. K., Kantor, D., & Finkelhor, G. (2009). Parent/child concordance about bullying
Houndoumadi, A., & Pateraki, L. (2001). Bullying and bullies in Greek elementary schools:
Hu, L.T., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure
Hunter, P. G., Schellenberg, E. G., & Schimmack, U. (2008). Mixed affective responses to
Hunt, C. (2007). The Effect of an Education Program on Attitudes and Beliefs about Bullying
and Bullying Behaviour in Junior Secondary School Students, Child and Adolescent
Innes-Ker, Å., & Niedenthal, P. M. (2002). Emotion concepts and emotional states in social
judgment and categorization. Journal of Personality and Social Psychology, 83, 804–
370
816.
Irwin, J. R. (2014). Development of an audiovisual speech perception app for children with
Bullying surveillance among youth: Uniform definitions for public health and
recommended data elements, version 1.0. Atlanta, GA: National Center for Injury
Prevention and Control, Centers for Disease Control and Prevention and U.S.
Department of Education.
Gorman, A. H., Dodge, K. A., &Bates, J. E. (2008). Friendships with peers who are low or
high in aggression as moderators of the link between peer victimization and declines
Jackson, D. L., Gillapsy, J. A., & Stephenson, P. (2009). Reporting Practices in Confirmatory
14(1), 6-23.
www.nspcc.org.uk/inform.
Jansen, P. W., Verlinden, M., Dommisse-van A. B., Mieloo, C., Ende, J. V. D., Veenstra, R.,
Verhulst,F. C., Jansen, W., & Tiemeier. H. (2012). Prevalence of bullying and
doi:10.1186/1471-2458-12-494.
371
Jansen, P. W., Verlinden, M., Dommisse-van, B. A., Mieloo, C., van der Ende, J., &
Jerman, P.G., Van der, L., & Agernbeaus, D. (2008). Heterogeneous inhibition process
Jones, S. E., Manstead, A., & Livingstone, A. (2012). Birds of a feather perpetrator together:
Jones, S. E., Manstead, A., & Livingstone, A. (2011). Ganging up or sticking together?
Jones, S. E., Haslam, S. A., York, L., & Ryan, M. K. (2008). Rotten apple or rotten barrel?
Joormann, J., Talbot, L., & Gotlib, I.H. (2007). Biased processing of emotional information
Junger-Tas, J., & Van Kesteren, J. (1999). Bullying and delinquency in a Dutch school
Junger, M. (1990). Intergroup bullying and racial harrassment in the Netherlands. Sociology
Juvonen, J. (2014). Bullying in school: The power of bullies and the plight of the
372
Juvonen, J., Wang, Y., & Espinoza, G. (2011). Bullying experiences and compromised
Juvonen, Jaana, Graham, Sandra, & Schuster, Mark. (2003). Bullying among young
adolescents:The strong, the weak, and the troubled. Pediatrics, 112(6), 1231-1237.
Juvonen, J., Nishina, A., & Graham, S. (2000). Peer harassment, psychological adjust-ment,
349-359.
Juvonen, J., & Wentzel, K. R. (1996). Social motivation: Understanding children's school
Kaukiainen, A., Bjorkqvist, K., Lagerspetz, K., Osterman, K., Salmivalli, C., & Rothberg, S.
(1999). The relationships between social intelligence, empathy, and three types of
Keefe, K., & Berndt, T. J. (1996). Relations of friendship quality to self-esteem in early
10.1177/0272431696016001007.
Kessler, H., Bayerl, P., Deighton, R. M., & Traue, H. C. (2002). Facially Expressed Emotion
Kelly, D., Liu, S., Rodger, H., Miellet, S., Ge, L., & Cal- dara, R. (2011). Developing
373
Khan, A. (Producer), Khan, A., Gupte, A., Madhvani, R. (Director). (2007). Taare Zameen
Khawar, R., Malik, F., Maqsood, S., Yasmin, T., & Habib, S. (2012). Age and Gender
Kim, Y. S., Koh, Y. J. & Leventhal, B. (2005). School Bullying and Suicidal Risk in
Kim, Y. S., Koh, Y-J. & Leventhal, B. L. (2004). Prevalence of school bullying in Korean
harassment through the Internet and other electronic means. Family & Community
Kline, R. B. (2011). Principles and practice of structural equation modeling. New York: The
Guilford Press.
Kline, R. B. (1998). Principles and practices of structural equation modeling. New York:
Kochanska, G., & Ladd, K., (1996). Inhibitory control in young children and its role in
Kokkinos, C. M., & Panayiotou, G. (2004). Associations with disruptive behavior disorders.
Koo, H., Kwak, K., & Smith, P. K. (2008). Victimization in Korean Schools: The Nature,
Koo, H. (2007). A time line of the evolution of school bullying in differing social contexts.
Kolb, B., Wilson, B., & Taylor, L. (1992). Developmental changes in the recognition and
comprehension of facial expression: Implications for frontal lobe function. Brain and
Kowalski, R. M., Limber, S. P., & Agatston, P. W. (2008). Cyber bullying: Bullying in the
Kristensen, S. M., & Smith, P. K. (2003). The use of coping strategies by Danish children
Kupersmidt, J. B., Griesler, P. C., DeRosier, M. E., Patterson, C. J., & Davis, P. W. (1995).
Childhood aggression and peer relations in the context of family and neighborhood
Kukučionytė, R., Drąsutė, V., & Drąsutis, S. (2012). The dissemination of best practices:
http://www.ceeol.com/aspx/getdocument.aspx?logid=5&id=295A0352-CF95-4C18-
9686-59250C8ED9B9.
375
Kujawa, R., Kucharczyk, D., Mamcarz, A., Żarski, D., &Targońska, K. (2011). Artificial
spawning of common tench Tinca tinca (Linnaeus, 1758), obtained from wild and
Kyllonen, P. C., Lipnevich, A. A., Burrus, J., & Roberts, R. D. (2010). Personality,
Kyriakides, L., Kaloyirou, C., & Lindsay, G. (2006). An analysis of the Revised Olweus
Ladd, G. W., Kochenderfer, B. J., & Coleman, C. C. (1996). Friendship quality as a predictor
of young children’s early school adjustment. Child Development, 67, 1103-1118. doi:
bu10.1111/j.1467-8624.1996.tb01785.x.
Laeheem, K., & Baka, D. (2009). Risk factors related to youths’ violence behaviors in the
Lamm, C., Lambert, L. L., Wolfe, J., Gray, J., Barber, A., & Edwards, G. (2014). improving
Socialization and Emotion Recognition for Children with Autism Using a Smartphone
Spectrum, (pp.128).
Lang, P., Bradley, M. M., & Cuthbert, B. N. (1997). Motivated attention:Affect, activation,
376
and action. In P. Lang, R. F. Simons, & M. Balaban (Eds) Attention and orienting:
Leach, C. W., Van Zomeren, M., Zebel, S., Vliek, M. L. W., Pennekamp, S. F., Doosje, B.,&
Leader, A. L., & Sethi, V. (1992). Root causes of strategic information systems planning
LeBlanc, A., Jin, Y. C., Obert, M., & Siivola, C. (1997). Effects of audience on music
Lee, W. (2011). Neural Substrates Of Intrinsic Motivation: Fmri Studies. University of Iowa,
America.
Lee, T., & Cornell, D. (2009).Concurrent Validity of the Olweus Bully/Victim Questionnaire.
Lee, T. M. C., Ng, E. H. H., Tang, S. W., & Chan, C. C. H. (2008). Effects of sad mood on
Leff, S. S., Kupersmidt, J. B., Patterson, C. J., & Power, T. J. (1999). Factors influencing
teacher identification of peer bullies and victims. School Psychology Review, 2, 505-
517.
Leppänen, J. M., Milders, M., Bell, J. S., Terriere, M., & Hietanen, J. K. (2004). Depression
133.
377
Leist, T., & Dadds, M. R. (2009). Adolescents’s ability to read different emotional faces
Lopes, P. N., Brackett, M. A., Nezlek, J., Schu¨tz, A., Sellin, I., & Salovey, P.
Limber, S. P., Olweus, D., & Wang, W. (2012). What we are learning about bullying: trends
in bullying over 5 years. Paper presented at the meeting of the International Bullying
Liu, J., & Graves, N. (2011). Childhood bullying: A review of constructs, concepts and
1446.2011.00972.x.
Liu, J., & Wuerker, A., (2005). Biological bases of aggressive and violent behavior-
ScienceDirect.com:http://www.sciencedirect.com.mimas.calstatela.edu/science.
Luby, J. L., Belden, Andy, Sullivan, Jill, & Spitznagel, Edward. (2007). Preschoolers'
young child self-report of symptoms. Child Psychiatry and Human Development, 38,
321-338. doi:10.1007/s10578-007-0063-8.
378
Lundqvist, L. O., Carlsson, F., Hilmersson, P., & Juslin, P. N. (2009). Emotional responses to
Luck, D. J., & Rubin, R. S. (1987). Marketing Research.Prentice Hall College Div.
Ma, X. (2002). Bullying in middle school: Individual and school characteristics of victims
Ma, L., Phelps, E., Lerner, J. V., & Lerner, R. M. (2009). Academic competence for
adolescents who bully and who are bullied: Findings from the 4-H study of positive
MacCallum, R. C., & Tucker, L. R. (1991). Representing Sources of Error in the Common-
Factor Model - Implications for Theory and Practice. Psychological Bulletin, 109(3),
502-511.
Mahady-Wilton, M. M., Craig, W. M., & Pepler, D. J. (2000). Emotional regulation and
Malik, F., & Shujja. S. (2013). Emotional Intelligence and Academic Achievement:
Malik, F., & Shujja, S. (2011). Initial validation and reliability analysis of social competence
scale for children: A cultural perspective. Journal of Behavioral Sciences, 2(1), 13-43.
379
Malik, F., Khawar, R., Chaudhry, H. R. & Humphreys, G. W. (2010). Emotion recognition
and duration of untreated first episode psychosis among patients in Pakistan. East
Markland, D. (2007). The golden rule is that there are no golden rules: A commentary on
Marsh, A. A., Ambady, N., & Kleck, R. E. (2005). The effects of fear and anger facial
Martin, L. L. & Clore, G. L. (2001). Theories of Mood and Cognition: A User's Guidebook.
Matsumoto, D. & Ekman, P. (1988). Japanese and Caucasian facial expressions of emotion
Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D.
Mayer, J. D., Allen, J., & Beauregard, K. (1995). Mood inductions for four specific moods:
Imagery,19, 133-150.
McCabe, S. B., Gotlib, I. H., & Martin, R. A. (2000). Cognitive vulnerability for depression:
McMahon, S. D., Felix, E.D., Petropoulos, L.A.N., & Halpert, J. (2009). Community
McNeilly-Choque, M., Hart, C. H., Robinson, C. C., Nelson, L. J., & Olsen, S. F. (1996).
Meerum Terwogt, M., & Olthof, T. (1989). Awareness and self-regulation of emotion in
MeerumTerwogt, M. (1986). Affective states and task performance in naïve and prompted
Meerum Terwogt, M., (1984). Colors and emotions: Preferences and combinations. Journal
Menesini E., Melan E., & Pignatti B., (2000). Interactional styles of bullies and victims
Menesini, E., Eslea, M., Smith, P. K., Genta, M. L., Fonzi, A., & Costabile, A.
Modecki, K. L., Minchin, J., Herbaugh, A. G., Guerra, N. G., & Runions, K. C. (2014).
Molcho, M., Craig, W., Due, P, Pickett, W., Harel-Fisch, Y., Overpeck, M., & the HBSC
1994– 2006: findings from Europe and North America. International Journal of
Monks, T. E., Smith, P. K., & Swettenham, M. (2003). The nature of school bullying and the
189-209.
Morita, Y., Soeda, H., Soeda, K., & Taki, M. (1999). Japan. In P. K. Smith, Y. Morita, J.
Junger-Tas, D. Olweus, R. Catalano, & P. Slee (Eds.), The nature of school bullying:
Mlisa, L.M., Ward, C.L., Flisher, A.J., & Lombard, C. (2008). Bullying at rural high schools
382
in the Eastern Cape Province, South Africa: Prevalence and risk and protective factors
Mynerd, H., & Joseph, S. (1997). Bull/victim problems and their association with
Nakamoto, J., & Schwartz, D. (2010). Is peer victimization associated with academic
National Crime Victimization Survey. (2009). Student Victimization in U.S. Schools. U.S.
Nesdale, W., Durkin, N., Maass, A., Kiesner, J., & Griffiths, J. A. (2008). Effects of Group
Neser, J. J., Ovens, M., van der Merwe, E., Morodi, R., & Ladikos, A. (2003). Peer
org.za/victismsp.pdf.
Nansel, T. R., Craig, W., Overpeck, M. D., Saluja, G., & Ruan, W. J. (2004). The Health
Nansel, T. Overpeck, M., Pilla, R.S., Ruan,W. J., Simmons-Morton, B., & Schmidt, P.
Association, 285,2094-2100.
Neil Tippett and Dieter Wolke. (2014). Socioeconomic Status and Bullying: A Meta-analysis.
Nelson, M. & deBacker, T. (2008). Achievement motivation in adolescents: the role of peer
climate and best friends. The Journal of Experimental Education, 76, 170- 189.
Nelson, C.A. (1987). Neural correlates of in- fants’ visual responsiveness to facial
Nickerson, A. B., Mele, D., & Princiotta, D. (2008). Attachment and empathy as predictors of
46, 687–703.
Nishina, A., & Juvonen, J. (2005). Daily reports of witnessing and ex- periencing peer
Noller, P., Feeney, J., & Peterson, C. C. (2001). Personal Relationships Across the Lifespan
Nordhagen, R., Nielsen, A., Stigum, H., & Köhler, L. (2005). Parental reported bullying
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory. New York: McGraw-
Hill.
Olafsen, R. N., & Viemerö, V. (2000). Bully/victim problems and coping with stress in
384
Oldenburg, B., Marijtje, D., Miranda, S., Gijs H., Rozemarijn P., Christina, S., &
33–44. doi:10.1007/s10802-013-9847-4.
O’Moore, M., & Kirkham, C. (2001). Self-esteem and its relationship to bullying behavior.
O’Moore, M., Kirkham, C., & Smith, M. (1997). Bullying behaviour in Irish schools: A
Osterman, K., l., Björkqvist, K., & Kaukiaine, A. (1998). Social intelligence − empathy
doi:10.1016/S1359-1789(98)00029-9.
Österman, P. Pil, F., Macduffie, J. P., Bjdrkqvist K., Kaukiainen, A. (1994). The
development of direct and indirect aggressive strategies in males and females. 176-
Norway.
Olweus, D., & Solberg, M. E. (2003). Prevalence estimation of school bullying with the
doi: 10.1002/ab.10047.
Olweus, D. (2001). Peer harassment: A critical analysis and some important issues, Peer
Olweus, D., & Endresen, I. M. (1998). The importance of sex-of-stimulus object: Age trends
Olweus, D. (1995). Annotation: Bullying at school: Basic facts and effects of a school based
Olweus, D. (1994). Bullying at School: Basic Facts and Effects of a School Based
Olweus, D. (1993). Bullying at school: What we know and what we can do. American
Aggression and Violence Throughout the Life Span. Sage. Newbury Park, CA.
Olweus, D. (1991). Bully/victim problems among schoolchildren: Basic facts and effects of a
Olweus, D. (1989) Bully/Victim Questionnaire for Students. (Circular No. 2001/10) Bergen,
Olweus, D. (1978). Aggression in the schools.Bullies and whipping boys. Washington, DC:
Nebraska Press.
Orobio de Castro, B., Merk, W., Koops, W., Veerman, J.W., & Bosc Orobio de Castro, B.,
Merk, W., Koops, W., Veerman, J.W., & Bosch, J.D. (2003). Emotions in social
information processing and their relations with reactive and information processing
and their relations with reactive and proactive aggression in referred proactive
Behavior, 5, 467-490.
Merk, W., Koops, W., Veerman, J.W., & Bosch, J.D. (2003). Emotions in social information
processing and their relations with reactive and information processing and their
490.
Pagliaccio, D., Luby, B., Gaffrey, M., Beldenb, A., Botteronb, K., Ian, H. G., Deanna, M.,
& Barcha, B. (2012). Anomalous functional brain activation following negative mood
Panayiotis, S., Anna, P., Charalambos, T., & Chrysostomos, L. (2010). Prevalence of
bullying among Cyprus elementary and high school students. International Journal of
Paquette, S., Peretz, I., & Belin, P. (2013) . The Musical Emotional Bursts": a validated set of
Psychology, 4, 509-510.
Parker, J. G., & Seal, J. (1996). Forming, losing, renewing, and replacing friendships:
Parker, J. G., & Asher, S. R. (1993). Peer relations and later personal adjustment: Are low-
Patterson, G. R., Reid, J. B., & Dishion, T. J. (1992). Antisocial boys. Eugene,
Oregon: Castalia.
Pepler, D., Jiang, D., Craig, W., & Connolly, J. (2008). Developmental trajectories of
Pellegrini, A. D., Long, J. D., Solberg, D., Roseth, C., DuPuis, D., Bohn, C., & Hickey, M.
(2011). Bullying and social status during school transitions. In S.R. Jimerson, S.M.
Pellegrini, A. D., & Burndt, M. (2001). Dominance in early adolescence boys: Affiliate
Peer Affiliation During the Transition From Primary School to Middle School,
Pellegrini, A., Bartini, M., Y., &Brooks, F. (1999). School Bullyes, Victims and Aggressive
Pepler, D. J., Craig, W. M., Ziegler, S. & Charach, A. (1993). A school-based anti-bullying
Perkins, K. A., Lerman, C., Grottenthaler, A., Ciccocioppo, M., Milanak, M., Conklin, C. A.,
Bergen, A. W., & Benowitz, N. L. (2008). Dopamine and opioid gene variants are
associated with increased smoking reward and reinforcement due to negative mood.
Pereira, B., Mendonc¸a, D., Netto, C., Almeida, A., Valente, L., & Smith, P. K. (1996). Facts
and gures of the rst survey on bullying in Portugese schools. Paper presented at the
Pérez-Edgar, K., Reeb-Sutherland, B. C., McDermott, J. M., White, L. K., Henderson, H. A.,
Degnan, K. A., Hane, A. A., Pine, D. S., & Fox, N. A. (2011). Attention biases to
threat link behavioral inhibition to social withdrawal over time in very young
Perren, S., & Alsaker, F. D. (2006). Social behavior and peer relationships of victims, bully-
47(1), 45-57.
Perry, D. G., Hodges, E. V. E., & Egan, S. K. (2001). Determinants of chronic victimization
by peers: A review and new model of family influence. Peer harassment in school:
The plight of the vulnerable and victimized, New York: Guilford Press.
389
Perry, G. D., Kusel, S. J., & Perry, C. L. (1988). Victims of peer aggression. Developmental
Peskin, M. F., Tortolero, S. R., & Markham, C. M. (2006). Bullying and victimization among
Pett, M., Lackey, N. & Sullivan, J. (2003).Making sense of factor analysis. Thousand Oaks:
Pine, D. S., Helfinstein, S. M., Bar-Haim, Y., Nelson, E., & Fox, N. A. (2009). Challenges in
Plybon, L. E., & Kliewer, W. (2001). Neighborhood types and externalizing behavior in
urban school-age children: Tests of direct, mediated and moderated effects. Journal of
Pollak, S. D., Cicchetti, D., Hornung, K., & Reed, A. (2000). Recognizing emotion in faces:
679–688.
Postigo, R. G., González, S., Montoya J., & Ordoñez G. (2013). Theoretical proposals in
doi.org/10.6018/analesps.29.2.148251 .
Poulou, M., & Norwich, B. (2002). Cognitive, emotional and behavioral responses to
2001 777-792.
Prinstein, P., Boergers, S., & Vernberg, E. M. (2001). Preliminary report of the Topeka
7(2), 183-203.
Raskauskas, J., & Stoltz, A. D. (2007). Involvement in traditional and electronic bullying
Reijntes, A., Albert, B., Kamphuis,J. H. ,Prinzie, K., Peter, P., & Michael, J.(2010). Peer
Renda, J., Vassallo, S., & Edwards, B. (2011). Bullying in early adolescence and its
Rennels, J. L., &Cummings, A. J. (2013). Sex differences in facial scanning: Similarities and
Reynolds, W.M. (2003). Bully Victimization: Reynolds Scales for Schools. San Antonio,
Rigby, K.(2007). Bullying in schools and what to do about it, ACER, Melbourne.
583-590.
Rigby, K. (2001). Stop the Bullying: A Handbook for Schools. Australian Council for
Rigby, K. (1997). Bullying in schools - and what to do about it. British Edition, London,
Jessica Kingsley.
Rivers, I. & Smith, P. K. (1994). Types of bullying behaviour and their correlates. Aggressive
Rigby, K., & Slee, P.T. (1991). Bullying among Australian school children: reported
Robazza, C., & Pellizzari, M., & Hanin, Y. (1994). Emotion self-regulation and athletic
379-404.
Robers, S., Kemp, J., and Truman, J. (2013). Indicators of School Crime and Safety: 2012
(NCES 2013-036/ NCJ 241446). National Center for Education Statistics, U.S.
Rodkin, P., & Berger, E. (2008). Bullies and victims in the peer ecology: Four questions for
Rodkin, P. C., & Hodges, E. V. (2003). Bullies and victims in the peer ecology: Four
32(3), 384-400.
Rodkin, P. C., Farmer, T. W., Pearl, R., & Van Acker, R. (2000). Heterogeneity of popular
Rosenfeld, L. B., Richman, J. M., & Bowen, G. L. (2000). Social support networks andschool
outcomes: The centrality of the teacher. Child and Adolescent Social Work Journal,
17(3), 205-226.
Ross, D. (1996). Review of literature, variety of strategies for guidance counselors and
Rothen, C. Head, J., Klineberg, E., & Stansfeld, S. (2011). Can social support protect bullied
from http://www.ijea.org/v7n9/.
Rubin, K. H., Bukowski, C. S. L., & Laursen,, N. A. (2009). Emotion regulation, parenting,
and the display of social reticence in preschoolers. Early Education and Development,
12, 97–115.
Rubin, K.H., Bukowski, W., & Parker, J. (1998). Peer interactions, relationships, and groups.
Rudatsikira, E., Muula, A. S., & Siziya, S. (2008). Variables associated with physical fighting
16-17.
Rueger, S. Y., & Jenkins, L. N. (2014). Effects of peer victimization on psychological and
393
doi: 10.1037/spq0000036.
Rutter, M. (1986). Child psychiatry: The interface between clinical and developmental
Saarni,C. (1990), emotional competence: how emotions and relationships become integrated.
Salmivalli, C. (2010). Bullying and the peer group: A review. Aggression and Violent
AB1>3.0.CO;2-T.
Salmivalli, C. (2002). Proactive and reactive aggression among school bullies, victims, and
Bullying as a group process: Participant roles and their relations to social status within
Salovey, P., & Mayer, J. D. (1990). Emotional Intelligemce. Imagination, Cognition and
Personality, 9, 185-211.
Scharphorn, A., & Laura, D. P. (2013). Between and Within- Child Associations of
Schäfer, M., Werner, N. E., & Crick, N. R. (2002). A comparison of two approaches to the
Scheithauer, H., Hayer, T., Petermann, F. & Jugert, G. (2006). Physical, verbal and relational
forms of bully‐ ing among students from Germany: Gender‐, age‐differences and
Schmid, P. C., Schmid Mast, M, Bombari, D., Mast, F. W. &Lobmaier, J.S. (2011). How
Schmid, P. C., &Schmid Mast, M. (2010). Mood effects on emotion recognition. Motivation
Schwartz, D., Gorman, A. H., Duong, M. T., & Nakamoto, J. (2008). Peer relationships and
Schwartz, D., Proctor L. J., & Chien, D. H. (2001). The aggressive victim of bullying:
Seals, D., & Young, J. (2003). Bullying and victimization: prevalence and relationship to
gender, grade level, ethnicity, self-esteem, and depression. Archives of Pediatrics &
Segun, P. (2004). Parental maltreatment and emotion dysregulation as risk factors for
30, 349-363.
Seixas, S. R., Coeelho, J. P., & Nicolas-Fischer, G. (2013). Bullies, Victims and Bully-Victims
http://www.fpce.up.pt/ciie/sites/default/files/05.SoniaRaqueletal.pdf.
Seixas, S. R., Coelho, J. P., & Nicolas-Fischer, G. (2013). Bullies, Victims And Bully-.
Selekman, A., & Vessey, J. A. (2004). Teasing and bullying experiences of middle school
Sethi, V., & King, W. R. (1991). Construct measurement in information systems research: An
Schmid, P. C., & Mast, M. (2010). Interpersonal behavior and social perception in a
hierarchy: The Interpersonal Power and Behaviour Model. European Review of Social
and relationally victimized boys and girls. Merrill-Palmer Quarterly, 53, 439-460.
84, 127-190.
the Mozart effect. In R. MacDonald, G. Kreutz, & L. Mitchell (Eds.), Music, health,
Schellenberg, E. G., Nakata, T., Hunter, P. G., & Tamoto, S. (2007). Exposure to music and
cognitive performance: Tests of children and adults. Psychology of Music, 35, 5–19.
Schuster, B. (1999). Outsiders at school: The prevalence of bullying and its relation with
Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being:
Sharp, C., Fonagy, P., & Goodyer, I. (2006). Imagining your child’s mind: Psychosocial
Sharp, S., Thompson, D., & Arora, T. (2000). How long before it hurts? School Psychology
Shaw, T., Dooley, J. J., Cross, D., Zubrick, S. R., & Waters, S. (2013). The Forms of
Bullying Scale (FBS): Validity and reliability estimates for a measure of bullying
Shear, M. K. (2011). Lifting the fog: Complicated grief and its treatment. Paper presented at
Sheridan, J. W., (2010). Short and long term effects of bullying- the victim & the bully.
Shujja, S., Atta, M., & Shujjat, S. (2014). Translation and Validation of Illinois Bullying
Scale for Pakistani Children and Adolescents, Pakistan Journal of Social and Clinical
Psychology, 9, 79-82.
Silver, M., & Oakes, P. (2001). Evaluation of a new computer intervention to teach people
Autism, 5, 229-316.
Simon-Morton, B., Crump, T. R., Haybie, M. O., & Saylor, P. H. (1998). Bullying
397
Smith, P. K. (2014). Understanding School Bullying: Its Nature and Prevention Strategies.
Smith, P. K., & Monks, C. P. (2008). Concepts of bullying: Developmental and cultural
Smith, P. K. (2004). Bullying: Recent Developments. Child and Adolescent Mental Health,
Smith, P. K., Cowie, H., Olafsson, R., & Liefooghe, A. (2002). Definitions of bullying: a
comparison of terms used, and age and sex differences, in a 14-country international
Smith, P. K., & Sharp, S. (1994). School bullying: Insights and perspectives. London:
Routledge.
Siris, K. & Osterman, K. (2004). Interrupting the cycle of bullying and victimization in the
27(2), 101-110.
Solberg, M. E., & Olweus, D. (2003).Prevalence estimation of school bullying with the
398
doi: 10.1002/ab.10047.
Solberg, M., Soderstrom, H. S., Salmon, G., James, A., Cassidy, E. L., & Javoloyes, M. A.
within a school for emotionally and behaviorally disturbed children. Clinical Child
Solomon, R. J., Solomon, S., Toumbourou, J. W., & Catalan, R. F. (2013). A cross-national
http://iussp.org/sites/default/files/event_call_for_papers/A%20cross-
national%20comparison%20of%20adolescent%20bullying%20victimization%20in%
20Mumbai%20India,%20Melbourne%20Australia%20and%20Seattle%20US.pdf
Sousa, V., & Rojjanasrirat, W. (2010). Translation, adaptation and validation of instruments
or scales for use in cross-cultural health care research: A clear and user friendly
Sourander, A., Helstelä, L., &Helenius, H. (2000). Persistence of bullying from childhood
to adolescence a longitudinal 8-year follow-up study. Child Abuse & Neglect, 24(7),
873–881.
Stagg, S., & Sheridan, D. (2010). Effectiveness of bullying and violence prevention
Stegge, H., Terwogt, M., & Koops, W. (1994). Positive and negative mood effects in
Stephenson, P., & Smith, D. (1989). Bullying in the junior school. In D. P. Tattum & D. A.
Stern, S. B., & Smith, C. A. (1995). Family processes and delinquency in an ecological
Storch, W. M., Brassard, D., &Masia-Warner, J. (2003). Identifying and targeting risk for
577- 582.
Sullivan, K. (2000). The Anti-Bullying Handbook. United Kingdom: Oxford University Press.
Sutton, J., Smith, P. K., & Swettenham, J. (1999). Bullying and theory of mind: A
critique of the social skills deficit” view of anti -social behaviour. Social
Sutton, J., & Colleagues, P. K. (1999). Bullying as a group process: An adaptation of the
Sutton, J. & Keogh, E. (2001). Social competition in school: Relationships with bullying,
443–456.
Sutton, J., Smith, P. K., & Swettenham, J. (2001). Social cognition and bullying: social
17(3), 74-78.
Suzuki, A., Hoshino, T., Shigemasu, K., & Kawamura, M. (2007). Decline or improvement?
75–84.
Swearer, S., Espelage, D., Vailancourt, T., & Hymel, S. (2010). What can be done about
400
39(1), 38–47.
Swearer, S. M., Espelage, D. L., Merrell, K.W., & Napolitano, S. A. (2009). Bullying
Boston, MA.
Swearer, S. M., Espelage, D. L., & Napolitano, S. A. (2009). Bullying Prevention and
Swearer, S. M., Peugh J., Espelage D. L., Siebecker A. B., Kingsbury W. L., &
Swearer, S. M. & Cary, P. T. (2003). Attitudes toward bullying in middle school youth: A
Syed, E. U., Hussein, S. A., Azam, S. I., Khan, A. G. (2009). Comparison of Urdu version of
the strengths and difficulties questionnaire (SDQ) and the Child Behaviour Checklist
Tabachnick, B.G., & Fidell, L. S. (2001). Using Multivariate Statistics. Boston: Allyn and
Bacon.
Thien, L. M., Razak, N. A., & Jamil, H.(2012). School of Educational Studies, Universiti
Taki, M. (2001). Relation among bullying, stress and stressor: A follow-up survey using
401
panel data and a comparative survey between japan and Australia. Japanese Society,
5, 25-40.
Tamir, M., & Robinson, M. D. (2007). The happy spotlight: Positive mood and selective
1124- 1136.
Thien, L. M., & Nordin, A. R. (2012). Academic Coping, Friendship Quality and Student
Thompson, M., & Cohen, L. (2005).When the bullied must adjust. Educutional Digest, 70(6),
16-19.
Toe, T., Tsai, T., & Yang, C. C. (2013). Applying Structural Equation Modeling (SEM) in
Tomal, D. R. (1998). A five-styles teacher discipline model. Paper presented at the annual
Tonks, J., Williams, H., Frampton, I., Yates, P.J., &Slater, A. (2007). Assessing emotion
Townsend, L., Flisher, A.J., Chikobvu, P, Lombard, C., & King, G. (2008). The relationship
402
between bullying behaviours and high school dropout in Cape Town, South Africa.
Ttofi, M., M., Farrington, D. P., & Lӧsel, F. (2012). School Bullying as a Predictor of
Ttofi, M. M., Farrington, D. P., Lösel. F., & loeber, B. B. (2011), Editorial: school bullying
and later criminal offending. Criminal Behaviour and Mental Health, 21, 77-79.
Troop-Gordon, W. & Quenelle, M. D., (2010). How does cognitive control reduce anger and
Tucker, L. R., & Lewis, C. (1973). A reliability coefficient for maximum likelihood factor
Topcu, C. & Erdur-Baker, O. (2012). Affective and cognitive empathy as mediators of gender
550-561. doi:10.1177/0143034312446882
Underwood, C. K., & Rosen, L. H. (2011). Gender and Bullying: Moving Beyond Mean
Underwood, M.K. (2003). Social aggression among girls. New York: Guilford
United Nations Children’s Fund (2012). Measuring and Monitoring Child Protection
Systems: Proposed Core Indicators for the East Asia and Pacific Region,
Unnever, D. J., & Cornell, D. G. (2004). Middle school victims of bullying: Who reports
U.S. Department of Health & Human Services. (2014). What is bullying: Bullying
Urberg, K. A., Degirmencioglu, S. D., & Pilgrim, C. (1997). Close friend and group influence
119.
Vaillancourt, T, McDougall, P., Hymel, S., Krygsman, A., Miller, J., Stiver, K., & Davis, C.
(2008). Bullying: Are researchers and children/youth talking about the same thing?
Van Kleef, G. A., Homan, A. C., Beersma, B., van Knippenberg, D., van Knippenberg, B., &
Vastfjall, D. (2002). Emotion induction through music: A review of the musical mood
Veenstra, R., Lindenberg, S., Munniksm, A., & Dijkstra, J. K. (2010). The Complex Relation
Attention to Status, Affection, and Sex Differences. Child development, 81(2) 480–
486.
404
Veenstra, R., Lindenberg, S., Oldehinkel, A. J., De Winter, A. F., Verhulst, F. C., & Ormel, J.
manuscript.
Véronneau, M-H., & Dishion, T. J. (2012). Friendship interactions as mediators of the link
between parental monitoring in high school and educational attainment by age 25.
Poster presented at the biennial meeting of the Society for Research in Child
Veronneau, M.-H., Vitaro, F., Pedersen, S., &Tremblay, R. E.(2008). Do peers contribute to
Vieno, A., Gini, G., & Santinello, M. (2011). Different forms of bullying and their
Wang, J., Iannotti, R., & Luk, J. W. (2011). Peer victimization and academic adjustment
Weiss, H. M., & Cropanzano, R. (1996). Affective events theory: A theoretical discussion of
Westermann, R., Spies, K., Stahl, G., & Hesse, F. W. (1996). Rela- tive effectiveness and
405
Whitney, I., & Smith, P. K. (1993). A survey of the nature and extent of bullying in primary
Wal, M. F., Wit, C. M., & Hirasing, R. A. (2003).Psychosocial Health among Young Victims
and Offenders of Direct and Indirect Bullying. Pediatrics, 111(6), 1312 -1317.
Values, Emotion Processes, and Social Competence. Social Development, 10(1), 59–
Wolfe, D. A., Scott, K., Reitzel-Jaffe, D., Wekerle, C., Grasley, C., &Straatman, A. L.
Wolke. D., William. E. C., Adrian, A. & Costello, E. J. (2013). Impact of Bullying in
Childhood on Adult Health, Wealth, Crime, and Social Outcomes. Association for
Wolke D, Woods S, Bloomfield L, & Karstadt L. (2000). The association between direct and
Woods, S., Wolke, D., Nowicki, S. & Hall, L. (2009) Emotion recognition abilities and
Woods, S., & Wolke, D. (2004). Direct and relational bullying among primary school
155.
Wong, D. S., Lok, D. P., Lo, T. W., & Ma, S. K. (2008). School bullying among Hong
406
605-618.
Ybarra, M.L. & Mitchell, K.J. (2004). Online aggressors, victims, and aggressor/victims: A
Ye, L., Ferdinando, H., Seppänen, T., & Alasaarela, E. (2014). Physical violence detection
http://http://www.hindawi.com/journals/aai/2014/740358/.
Yong, T., Hua, M., & Mei, G. (2007). Empirical study on impact of logistics operations
Shanghai, China.
Zeller, M. H., Vannatta, K., Schafer, J., & Noll, R. B. (2003). Behavioral reputation: A
Zindi, F. (1994). Bullying at boarding school: A Zimbabwe study. Research in Education, 51,
23-32.
407
Appendix A
(Sample Items from OBVQ and FQS Urdu Version)
408
409
410
Appendix B
(CFA of OBVQ)
411
412
RMR, GFI
Model RMR GFI AGFI PGFI
Default model .047 .949 .933 .719
Saturated model .000 1.000
Independence model .330 .424 .347 .374
Baseline Comparisons
NFI RFI IFI TLI
Model CFI
Delta1 rho1 Delta2 rho2
Default model .915 .901 .939 .929 .939
Saturated model 1.000 1.000 1.000
Independence model .000 .000 .000 .000 .000
Parsimony-Adjusted Measures
Model PRATIO PNFI PCFI
Default model .858 .785 .806
Saturated model .000 .000 .000
Independence model 1.000 .000 .000
RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .052 .046 .059 .252
Independence model .197 .192 .202 .000
AIC
ECVI
Model ECVI LO 90 HI 90 MECVI
Default model .490 .427 .563 .492
415
HOELTER
HOELTER HOELTER
Model
.05 .01
Default model 312 341
Independence model 31 34
416
Appendix C
(CFA of FQS)
417
CMIN
Model NPAR CMIN DF P CMIN/DF
Default model 46 358.017 164 .000 2.183
Saturated model 210 .000 0
Independence model 20 2625.869 190 .000 13.820
RMR, GFI
Model RMR GFI AGFI PGFI
Default model .067 .949 .934 .741
Saturated model .000 1.000
Independence model .289 .543 .495 .491
Baseline Comparisons
NFI RFI IFI TLI
Model CFI
Delta1 rho1 Delta2 rho2
Default model .864 .842 .921 .908 .920
Saturated model 1.000 1.000 1.000
Independence model .000 .000 .000 .000 .000
Parsimony-Adjusted Measures
Model PRATIO PNFI PCFI
Default model .863 .745 .794
Saturated model .000 .000 .000
Independence model 1.000 .000 .000
421
NCP
Model NCP LO 90 HI 90
Default model 194.017 143.191 252.585
Saturated model .000 .000 .000
Independence model 2435.869 2273.995 2605.104
FMIN
Model FMIN F0 LO 90 HI 90
Default model .534 .289 .213 .376
Saturated model .000 .000 .000 .000
Independence model 3.913 3.630 3.389 3.882
RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .042 .036 .048 .988
Independence model .138 .134 .143 .000
AIC
Model AIC BCC BIC CAIC
Default model 450.017 452.989 657.489 703.489
Saturated model 420.000 433.569 1367.154 1577.154
Independence model 2665.869 2667.161 2756.074 2776.074
ECVI
Model ECVI LO 90 HI 90 MECVI
Default model .671 .595 .758 .675
Saturated model .626 .626 .626 .646
Independence model 3.973 3.732 4.225 3.975
HOELTER
HOELTER HOELTER
Model
.05 .01
Default model 366 392
Independence model 58 61
422
RMR, GFI
Model RMR GFI AGFI PGFI
Default model .060 .963 .951 .721
Saturated model .000 1.000
Independence model .287 .571 .520 .511
Baseline Comparisons
NFI RFI IFI TLI
Model CFI
Delta1 rho1 Delta2 rho2
Default model .901 .881 .955 .945 .954
Saturated model 1.000 1.000 1.000
Independence model .000 .000 .000 .000 .000
Parsimony-Adjusted Measures
Model PRATIO PNFI PCFI
Default model .837 .754 .798
Saturated model .000 .000 .000
Independence model 1.000 .000 .000
NCP
Model NCP LO 90 HI 90
Default model 96.130 58.381 141.740
Saturated model .000 .000 .000
Independence model 2106.918 1956.834 2264.374
FMIN
Model FMIN F0 LO 90 HI 90
Default model .334 .143 .087 .211
Saturated model .000 .000 .000 .000
Independence model 3.368 3.140 2.916 3.375
424
RMSEA
Model RMSEA LO 90 HI 90 PCLOSE
Default model .033 .026 .041 1.000
Independence model .143 .138 .149 .000
AIC
Model AIC BCC BIC CAIC
Default model 310.130 312.637 504.072 547.072
Saturated model 342.000 351.966 1113.254 1284.254
Independence model 2295.918 2296.967 2377.103 2395.103
ECVI
Model ECVI LO 90 HI 90 MECVI
Default model .462 .406 .530 .466
Saturated model .510 .510 .510 .525
Independence model 3.422 3.198 3.656 3.423
HOELTER
HOELTER HOELTER
Model
.05 .01
Default model 466 504
Independence model 55 59
425
Appendix D
(Demographic Information Sheet)
426
427
Appendix E
(Permission from Schools, Information for Parents and Consent Letters)
428
429
430
431
Appendix F
(Sample Emotive Pictures for Mood Induction)
432
Appendix G
(Plagiarism Report)