Mauritian Childrens Moral Judgments About Bullying Psychology Essay

Published Date: 23 Mar 2015

The current study investigated whether young Mauritian children's moral judgments about bullying of siblings and peers on different activities and scenarios on the basis of gender and ethnicity would vary according to their grade (age), ethnicity (Hindus, Muslims and General-populations which include Sino-Mauritians and Christians) level and the context in which the bullying occurred. Individual interviews about bullying in several different contexts were conducted with 120 children from grade 2, 4 and 6. They were expected to reject any type of bullying: direct, relational and cyberbullying, by using judgments based on moral reasoning and/or social conventional reasonings. There were no significant relationships between ethnicity, grade or gender of participant with bullying. Children were found to make more use of moral reasoning to justify whether they support or condemn bullying practices. Also, bullies made less use of reasoning to disapprove of bulling. The findings are discussed in relation to the Mauritian culture and intergroup relations among peers in schools. Limitations of the study were discussed and recommendations for future research were suggested.

Keywords: Moral reasoning; moral judgment; social-conventional reasoning; bullying; direct bullying; relational bullying; cyber-bullying.

CHAPTER I

INTRODUCTION AND LITERATURE REVIEW

Introduction

This chapter is mostly on the areas of literature related to the research. Firstly, a brief introduction of primary education in Mauritius, followed by a description on bullying (including definition, prevalence, status and bullying on different target group) and finally the relation of moral reasoning based on different theories with bullying (in context of sibling, peer, gender and ethnicity) will be presented. Last of all, an outline of the aims and hypotheses will be explored.

No country including Mauritius can say it is "a model of caring" in its public or private educational institutions (Smith, 2003). Research done worldwide show that school violence is an international health apprehension (Smith, 2003; Smith, Olweus, et al., 1999). Children can be involved in violence in any of their surroundings: home with families, school, day-care and other communities.

Schools in Mauritius are microcosms of its society; it comprises of children of all religions, socio-economic status and social background. Education on the island is managed by the Ministry of Education and Human Resources. It controls the administration of state schools and also has a supervisory role on private schools. The Ministry ensures that all children, as they reach 5 years of age, are enrolled in one of the 305 primary schools scattered on the island. It also sees to it that all children, irrespective of their various differences, enjoy equal rights and facilities. Each and every child goes through a fair admission process, whereby the school allocated is the one nearest to his/her residence, studies the same curriculum and is provided with free education, free transport, free school books, bread and appropriate recreational facilities.

Just like any interesting psychological notion, there is no common agreement on how to define 'bullying'. However, there is a consensus with researchers in the domain that it is a form of violent and aggressive behaviour which involve the following (Mills & Duck, 2000):

o Behaviour that can physically and/or psychologically harm/hurt someone else

o Having the intention to do so

o Repetitive

o Where there is power imbalance and the victim is not able to retaliate and/or defend himself/herself.

Early research has identified three types bulling: physical, verbal or relational. The first two are also known as direct bullying since it is visible and they embrace direct aggressive acts such as kicking, pushing, name-calling, teasing, humiliating, degrading and other types of direct verbal abuse (Bjorkqvist, 1994). Crick and Grotpeter (1995) defined relational bullying as causing harm on peers such as exclusion from games, spreading rumours, which can destroy peer relationships. Relational bullying is not easily distinguishable. The wild proliferation of new technology in recent years has given students an increasing access to electronic communication such as mobile phones and computers. This has contributed to the rise of cyberbullying: the use of technology to bully others (Li, 2007). Willard (2004, as cited in Li, 2007), in her book, acknowledged some common types of cyber-bullying: flaming (sending of vulgar and nasty messages), online harassment, cyberstalking, denigration (sending of destructive false statements to other people), masquerade, outing (publishing or sending humiliating, sensitive or private information) and exclusion (Rejecting someone from an online group). These types of bullying are important in relation to age and sex differences. Lagerspetz and Bjokqvist (1994) have studied Finnish children between the age of eight to eighteen years old. Despite that no differences were found in verbal bullying between boys and girls, boys (excluding young ones) were found to display more physical bullying while girls (excluding young ones) exhibited more indirect aggression (Monks & Smith, 2000).

Aggressive behaviours including bullying can have short and long term negative concerns such as suicide tendencies, isolation, depression, conduct disorders, delinquency, for the mental health and adjustment for victims (Coie & Dodge, 1998; Hawker & Boulton, 2000; Rubin, Bukowski, & Parker, 2006). The researchers in the field will be more able to plan effective intervention programs to alleviate the bullying problem and treat the consequences bullying have on mental health if they have a complete understanding of the factors involved in a child's decision to bully.

Peer relations. One of the developing tasks of children is to cultivate positive relation with peer - a relation that comprises of mutual understanding, acceptance, fun, self- improvement and also dominance. But since conflict is an inevitable component of human relations, relation between peers can occasionally turn into act of aggression (Slee & Rigby, 1998). Both the nature and quality of school peer group is inviting substantial research attention. This is because school peer group is an essential source of nurturance and support. However, any disruption in the relation can affect a student's physical and psychological welfare and nature of learning experiences (Bennet & Derevensky, 1995). It was found that there is a negative relation between the trend of being ill-treated and victimised and positive self-assessments of the number of friends and popularity (Slee & Rigby, 1993)

Sibling relations. Bullying between siblings is becoming a major source of concern for many families (Wolke and Samara, 2009). Children young enough for schools get tangled in physical aggression and continuous teasing (Dunn & Munn, 1985). Relationships between siblings are known to aid children in developing social skills, providing emotional support (Stormshak, Bellati, & Bierman, 1996) and act as a cushion in family hardships (Jenkins & Smith, 1990) such as violent parents or marital fights. Despite of the negative influence of sibling bullying and its dangerously obvious high rates of victimization, it has somewhat been overlooked in previous study, most probably because it occurs so frequently in most families (Goodwin & Roscoe, 1990). In a sample of 7th and 8th graders, Duncan (1999) found that about 30 % of the pupils were physically bullied by their siblings while 40% agree of bullying their siblings.

Bullying and Gender. Hallinan and Williams (1989) claimed that we can expect the prevalence of the negative effect on young children's friendships to surpass race and other background physiognomies. Though rare, cross-gender friendships do occur during elementary schools. According to Hallinan and colleague's research, friendship that went beyond ethnic and racial obstacles was from the same gender. During pre-school years, Fabes, Martin and Hanish (2003) found that female children looked for other girls as playmates because they prefer to be involved in unobtrusive activities which include cooperative roles while male children preferred to play with a large group of boys who share a need in active thrilling games like run, kick, roar, race, play-fight and knock down. Also, Theimer, Killen and Stangore (2001) studied the evaluations of preschool-aged children. They established that compared to the girls, the boys were less likely to evaluate exclusion of a child as negative.

Bullying and Ethnicity. One of the main reasons that children give for bullying others is that they are not similar either in the way they behave or in appearance (Terasahjo & Salmivalli, 2003). Race or ethnicity is one discernible way children can differ. Ethnicity is used to denote to variances in ancestry and lineage for which the primary cues are cultural (Slee & Rigby, 1998). Compared to race which comprises of a wide range of skin colour and other physical characteristics, ethnicity includes a catholic array of dialects, beliefs and customs. Past research (Wolke et al., 2001; Hanish & Guerra, 2000) have shown that there is a high chance for children to be victimised based on their race or ethnicity. Hanish and Guerra (2000) found evidence that minority groups' likelihood of being bullied were higher than the ethnic majority groups. In their longitudinal study, they found that irrespective of grade (from 1st grade to 6th grade), there were a higher probability of victimisation of European - American and African - American pupils by peers than Hispanic pupils. Furthermore, the composition of schools in terms of race and ethnicity had an influence on European-American and African-American children. European-American children who attended homogeneous schools were less likely to experience peer victimization than those who attended heterogeneous schools. Also, African-American pupils who joined ethnically heterogeneous schools decreased their likelihood to be victims (Hanish & Guerra, 2000). However, there are many studies which did not find any differences. Same results were obtained in two different investigations by Moran, Smith, Thompson and Whitney (1993) and Boulton (1995a) with Asian and White children for age, gender and grade. There were no significant differences in in terms of having friends or peer ratings of students as victims or bullies. Nevertheless, a substantial number of Asian children reported racist name calling by children from other races. Siann and his associates (1994) studied the impact ethnic group had in bullying in schools in Scotland and England. Again, there were no significant differences in the levels of bullying and victimization between the ethnic groups. Still, when asked whether minority groups are bullied more than the majority, 75.4 % from the minority ethnic group said yes compared to a 49.6% from the majority. Likewise, in a multi ethnic/racial sample of Australian children from seventh to tenth grade, Nguy and Hunt (2004) observed that there were no variances in the frequency of bullying or being bullied, in the attitudes towards the outgroups and no differences in the reasons given by the majority and minority.

In this study, the researcher has focused especially gender and ethnicity. According to Aboud (1992), gender and ethnicity are the most important social group membership categories to develop in a child. Also, children more usually exclude someone from their peer group based on its gender (Maccoby, 1988 as cited in Killen & Stangor, 2001) and race/ethnicity (Aboud, 1992).

Moral Reasoning in children

The psychoanalytic theory and the social learning theory look at moral development as a procedure of embracing social norms. If moral code is not internalised and shared and empathy is not refined through inductive discipline, people would then neglect the rights of one another every time there are conflicts with their desires and would transgress whenever other people could not observe their behavior. Researchers even though they give support to different theories ask themselves the same question: When do children become capable of moral reasoning? For the last few years, with the cropping up of new methods and theories, there have been drastic changes (from global stage theory and more towards the domain-specific models) in the research on the moral judgment of children (Killen, 2007)

Global Stage Model. Piaget's (1932-1965) early works on moral judgments of children were inspired by the cognitive-developmental perspective which is based on how children actively attend to and inter-relate numerous perceptions on situations in which social conflicts arise and thus are able to make moral judgments on the basis of notions they construct about justice and fairness (Berk, 2006). He identified two stages for the understanding of morals: the heteronomous and autonomous morality. In the first stage children look at moral rules as fixed, pragmatic and set by authority figures while in the second one, children's moral reasoning is focused on fairness on ideal mutuality and they realise that rules can be flexible (Berk, 2006).

Lawrence Kohlberg's global stage theory (three levels and 6 stages- see Figure 1) is an extension of Piaget's work. According to him, moral development is a gradual process that outspreads childhood to adolescence and adulthood.

Figure 1 : Kolhberg's Stages of Moral Development

Moral Domain Theory. On the other side, we have the Moral Domain Theory that argued that children have an undeveloped grip of justice as from age three. Elliot Turiel (1998) identified three domains of action: moral imperatives (based on how we should to treat one another and protecting people's equal rights and welfare), social conventions (customs designed mainly by consensus and regulations determined in order for social institution function smoothly) and psychological ( personal choice of individual with an understanding of self, and others and believe that everyone has his/her own autonomy and independence (Berk, 2006; Ardila-Rey & Killen, 2001).

Smetana (2006) has recounted the changes with age that past research had found within the moral domain. Children determine that moral transgressions are immoral in any context by dynamically evaluating people's daily social practices and emotive responses (Berk, 2006).

Social Information Processing Model (SIP). The approach of SIP (Crick & Dodge, 1994) helps us to understand meaningfully the development of aggressive behaviour (Orobio de Castro, 2004). The model suggests that to be able to respond in an appropriate manner to social cues, individuals have to process information in an orderly manner. First of all, a child will initially observe and try to understand the various hints from a social interaction to make sense of what happened and why it did. Then, guided by what he understood and relying on his previous exposure to similar interactions, he will opt for the simplest and most fitting objectives (Arsenio & Lemerise, 2004). Afterwards, the child will think of a range of likely reactions for the said situation and will weigh them according to his ability to produce those reactions, taking into account the different outcomes of what he intends to do. Eventually, he will reproduce what he has chosen. Depending on the conditions however, some infants might not go through all these steps before responding. For example, in circumstances where the child is very excited or is experiencing a strong emotion, he will probably act on impulse (Crick & Dodge, 1994). Indeed, angry children are known to express unfriendly characteristics and aims (Koops, & Veerman, 2003).

The circumstance, the person's ability to process information and something called a "database" are all involved in how someone goes through the phases in the model. The database is where all the past experiences are held either as links, memories or organised patterns which are then needed on every information- processing phase. As the database gets updated through experiences, social information- processing tends to take place at a higher rate and more efficiently (Crick & Dodge, 1994; Dodge, 1993). Several studies relating to SIP in children who are either bullied or incited by their friends have been carried out and these show that hostile attitudes in children is associated with an abnormal encoding, interaction aims, production of a reaction and database schemata (e.g., Dodge, 1980, 1993; Dodge, Lochman, Harnish, Bates, & Pettit, 1997; Lochman & Dodge, 1998). Less reliable results have been obtained for representation. This has mainly been looked at by making sense of the motives behind other people's behaviour (also called "hostile attribution bias"). A freshly conducted meta-analysis has shown that the attribution of motives and hostile behaviour are closely connected. However, all the research done on this subject have yielded very different results (Orobio de Castro, Veerman, Koops, Bosch, & Monshouwer, 2002).

SIP patterns have been found to be the link between early risks factors and future aggressive behaviour according to longitudinal studies (e.g., Pettit, Dodge, & Brown, 1988). Therefore, SIP by aggressive children is a goal for cognitive behavioural interventions to decrease behaviour problems. Interventions including SIP change (Hudley & Graham, 1993; Lochman & Wells, 2002) are quite good (Kazdin, 2003) as mediation analyses have shown that alterations in aggressive behaviour after these measures are in effect due to the changes in SIP (Lochman & Wells, 2002).

Integration of the models. The role of moral intentions and evaluations is a possible area of integration between domain and SIP models. In domain theory (and other cognitive moral models), the moral (or non-moral) type of the act is determined mainly by the children's judgement about whether the act was harmful or not. Unlike the other model, SIP argues that it is the second step (among the six steps), which is interpretation of cues, whereby children make intentional attributions about how other people behaved. An example would be when a child experiences a negative social outcome, such as getting hit by a ball or not being allowed to take part in a game. In this instance, the child responds only after analysing the intention of others e.g. was the ball thrown at him/her deliberately? Is he or she sitting out the game since the maximum number of players had been reached or is it because others wanted to be rude?

Morally, during the early stage of SIP research, relevant intentions were a main area of interest. During his first research into social reasoning of aggressive and non-aggressive boys, Dodge (1980) found that the social response of the boys were largely subjective to their perspective of the instigators' intentions leading to a negative outcome. The difference between the responses of the two groups appeared in situations where the provocateurs' intentions were unclear. ''Aggressive children responded as if the peer had acted with a hostile intent. Nonaggressive subjects responded as if the peer had reacted with benign intent'' (p. 162). There are various studies on intent judgements that have been carried out which back the idea that both group of children consider the equality and moral legitimacy of others' actions to respond (Coie & Dodge, 1998). Retaliation is considered justified in the perspective of a child if he/she judges that the harmful action of a peer was both aggressive and unfair and it was created purposely (Astor, 1994). Therefore it is fair to say that both aggressive and non-aggressive children can be considered to share a fundamental moral value; it is not right for someone else to harm the child intentionally. But following SIP research suggested that this response may only be suitable to a subtype of aggression. Important motivational core which are lacking in cognitive theories of aggression are brought forward by both the SIP and domain models. For example, according to the SIP model, in the framework of socially challenging interpersonal circumstances, socially proficient children are more interested to relational goals such as preserving friendships whereas aggressive children have a tendency for instrumental goals such as being in control of an object or a situation. In both theories, it is implied that most children care about their relationships with others and are affected by the negative outcomes of aggressive and immoral behaviours experienced by others.

Intergroup Contact Theory. There is panoply of studies that show that increasing intergroup contact can be associated with a reduction in ingroup bias in children and adolescents. From the age three to ten years, children are opened to a wider range of social influences which include school and media, meaning that over time contacts with people from other groups (for example nationalities, ethnic groups, religions) will rise (Cameron et al.,2011). If they have regular contact towards the outgroup members, they will be more exposed to their lifestyles and behaviours which can help to promote tolerant attitudes. On the other side, children may develop adverse stereotypical opinions if they have restricted contact and hence a lack of knowledge of how the outgroups function (Ellison & Powers, 1994). Still, Allport (1954, as cited in Pettigrew & Tropp, 2006) postulate that fruitful intergroup contact can only occur in four conditions: equal status between the groups; collaboration to achieve common goals; no inter-group competition; and the support from authorities, regulations, or laws.

In one of the first longitudinal studies, for two years Stephan and Rosenfield (1978) observed White American children's attitudes toward ethnic minority groups in school settings. More positive intergroup attitudes were found to result from an increase in inter-ethnic contact (as cited in Binder et al., 2009). Again in America, Brown (1986) showed a decrease in prejudice between black and white students. A scheme was applied where all students were given equal opportunities and little performance-streaming, and cooperation needed between students was increased (as cited in Hope, n.d.). Likewise, supportive research for Allport's conditions was found in a study by Islam & Hewstone (1993) (as cited in Hope, n.d.). They found that less prejudicial attitudes towards students of Islamic and Hinduism beliefs were linked to the quality and quantity of interaction between them. Moreover, Wagner et al. (2003) confirmed Allport's theory in their research with East and West German school students with a mean age of 15.5 years. The central role of intergroup friendship and the descriptive power of intergroup contact were found to elucidate the East-West variance in prejudice.

In another longitudinal study, Feddes, Noack and Rutland (2009) observed Turkish (minority) and German (majority) children's intergroup attitudes among each other and their friendship. The latters had already spent at least 2 years together. The researchers found that over time, extended cross-ethnic friendship resulted in more negative outgroup evaluations than direct one. Pettigrew (1998) said that the development of friendship between the ethnics and the tolerant nature of the children were possible because the conditions echoed those of Allport (1954): support from school authorities and the maintenance of equal status by acknowledging of the Turkish language through language courses. Similarly, European-American majority children who made regular interactions with the minorities did not make ingroup bias, unlike the majorities who made little connection with the minorities as observed by McGlothlin and Killen (2006).

More recently, Crystal and colleagues (2008) wanted to find out how school children, who had various and different types of opportunities to make intergroup interaction, assess exclusion of mixed-ethnicity children. Exclusion by race was judged as immoral by both the minorities and the majorities, showing that having inter-racial friends alleviate the chance of racial exclusion (as cited in Ruck, Crystal et al., 2011). Likewise, Killen et al. (2010) studied how European-American children and adolescents assess, judge the exclusion of the minorities based on their race and how they explained the racial stereotypes. Results suggested that compared with the majority children (European-American) with lower levels of intergroup contact, those who made higher intergroup contacts were more likely to view exclusion of races as inappropriate and incorrect, to oppose and forbid stereotypes.

In 2003, in an attempt to study the link between moral reasoning and bullying, Menesini, Codecasa, Benelli and Cowie looked at how 10- and 13- year olds made emotional attributions and justifications in a fabricated study of a bully. Results showed that those who bully qualified the malefactor as proud and indifferent more frequently than those who were victims or not involved. Similarly, in a sample of 9- year olds, Perren, Gutzwiller-Helfenfinger, Malti and Hymel (2011) found that bullies were more morally detached and justified less responsibly.

As mentioned earlier, race and ethnicity do influence assessments of bullying. In a survey with 265 European-American students from 6th and 9th grades, Margie (2007) studied the relationship between children's personal involvements with bullying, their social reasoning and their Social Information Processing (SIP) and how race/ethnicity can have an impact. Their judgments, how the children justify themselves, their goals, attributions, their choices of response in both same (European-American) and cross race (African-American) peer interactions were assessed along with their experiences and encounters with bullying. Participants with frequent bullying experience assessed the actions of bullies as less wrong, tended to blame the victim for the action of the bully and also paid less attention to the feelings of the victim. They were more hostile and less confident in their answers and preferred less personal and more aggressive goals for the latter. Moreover, in reviews of Naito, Lin and Gielen (2001), children and adolescents from Eastern Asian countries tend to do moral reasoning more empathy-based with pro-social intents, a good understanding of normative beliefs, personal endorsement and feelings of culpability compared to those in Western Asian countries. The latter were more likely to reason in an egocentric and self-indulgent perspective. However, Matsui (1997) could not replicate the same results. He found that Japanese students (Eastern Asian country) were twice likely to give reasons full of empathy to help others (e.g., "I would feel sorry for the other person and want to help") and less likely to voice reasons on their apprehensions about norms and rules as compared to students in the United States, South Korea, Turkey and China. In addition, Naito and colleagues (2001) found that in overall, the Japanese students are more likely to show a Western 'individualistic' reasoning compared to the other Eastern Asian countries who tend to show an Eastern "collectivistic" orientation.

In spite of having a robust link from theories (Killen & Nucci, 1995; Smetana, 2006; Wolke, Woods, Stanford, & Schulz, 2001) that bullying lies perfectly in the moral domain and is prone by social reasoning (Arsenio & Lemerise, 2004), insufficient studies have looked at the moral perspective of a child in the evaluation of bullying. Additionally, we do not have any empirical evidence of how moral reasoning is related to experiences with, even though abstract work advocates that the SIP influenced both of them (Arsenio & Lemerise, 2004).

However, past studies had some restrictions and limitations. To start with, they evaluate how bullies process social information only in provocative circumstances; that is they only consider what bullies would do, were they the victims. How do children who bully others process social information when they are bullying? For example, what are the social goals of the bully set against that of the bullied in a bullying situation? Also, these studies differentiate between groups of children based on whether they are the bullies or the victims. Whilst this approach takes into account those who bullies or are bullied to a large extent, it fails to capture the encounters of most children. Several of them who bully or are bullied in smaller frequency than those categorised as 'bullies' or 'victims in these studies'. These experiences are very likely to have impact on their reasoning and processing regarding bullying circumstances. For that reason, it is essential to look into a range of experiences with bullies and/or victims. Only a few researches have studied the social reasoning of violent children using the social cognitive domain theory. In general, these studies dealt with the social reasoning of young offenders and children having behavioural disorder, who demonstrated a variety of antisocial behaviours, of which, only one could be aggression. More directly linked to bullying, Astor (1994) who made the comparison between the social reasoning of violent children and non-aggressive ones, found that the two groups of children uses different aspects of the moral domain to justify moral situations. Definitely, physically violent children unlike the non-violents, were more tended to think that physical retaliation was proper because it signifies fairness contrary to moral because of the harm inflicted on others (Astor, 1994). This research will focus on normal children, not just aggressive children with known delinquent behaviours and will look at the differences among normal children based on their experiences as victims and bullies. Also, preceding investigations on bullying focused mostly on physical aggression and exclusion, a non-physical form of aggression (Killen et al., 2002; Phinney & Cobb, 1996). There are several other forms of bullying such as teasing, taking others' belongings, sending nasty messages, spreading rumours (and many other physical, verbal, relational and cyber bullying activities) that need to be examined.

This research is definitely a primary step to better understand the social reasoning of violent youngsters. The three main goals of the study are firstly to examine personal bullying experiences of the pupils based on their gender, ethnicity and grade, secondly to look at the relation between different bullying situations and children's social reasoning and lastly to study the effect of the gender and ethnicity of the target on the children's social reasoning. Altogether, this study delivers an opening for the analysis of bullying/victimization of siblings and most specifically peers within the ethnic and gender perspective. Some research has previously studied how the context ethnicity and gender are related to children's judgments (Lawrence, 1991; Margie et al., 2005; McGlothlin & Killen, 2006; McGlothlin et al., 2005; Moller & Tenenbaum, 2011). Nonetheless, this study will address the limitations and boundaries of previous works. Lastly, preceding works on moral reasoning do not look at the features and characteristics of social cognition that are assumed to have a direct link on the behaviours of children in a social situation (similar to the steps of SIP). It will be better for the experts in the field to understand the prognoses of bullying behavior if they get an improved version of the reasoning and mental processes of children that follows cross-ethnic and cross-gender exchanges including bullying. The current study moved beyond social exclusion and physical bullying and addresses other different types of victimization. Additionally, this study surveyed developmental (gradual development of children from grade 2 to grade 6) and gender differences between reasoning and bullying/victimization. The children were presented with hypothetical circumstances in which a child (sibling or peer) was being bullied or victimised. Gender, ethnic or peer group were the physiognomies of the target pupil that were varied as pertinent to the domain.

There were five sets of hypotheses in the study. Firstly, just like research on the social-cognitive theory that children including aggressive ones, generally condone maltreatment towards others (Killen & Nucci, 1995 and studies by Killen et al., (2000) and Moller and Tenenbaum (2011), we expected the children to be sensitive to the scenarios of the vignettes. In other words, we expected the pupils to judge bullying based on ethnicity as less acceptable than bullying based on gender and bullying on siblings as the most acceptable ( since previous literature such as Goodwin & Roscoe (1990) suggested it as very common). Secondly, guided by Abrams et al. (2009), Killen et al. (2000) and Moller et al. (2011), we anticipated some age differences. Older participants (Grade 6) were expected to view bullying as less acceptable as would younger ones (Grade 2 and Grade 4). Thirdly, we foretold that children will make more use of moral reasoning in the vignettes comprising of ethnicity of target followed by gender of target (Moller and Tenenbaum, 2011) and would use more moral reasoning than social-conventional reasoning in vignettes involving ethnic bullying than in vignettes on gender bullying. In addition, it was put forward that on a whole, children would consider bullying as wrong but more specifically, children who were found to be bullies would be less likely to provide reasoning to disapprove of bullies' actions (Menesini et al., 2003). Finally, it was postulated that those who give support to bullying will make more use of stereotype justifications (Margie, 2007).

Chapter 2

METHODOLOGY

This chapter is devoted to the methodology applied to test these research hypotheses. Particularly, it discusses the research design, sampling procedure, and the data analysis plan for the empirical study.

Sample Profile (Research Participants)

One hundred and twenty primary children aged 6 to 12 years old (M = 9.31, SD= 1.61) from eight different public schools located in socioeconomically varied areas in Mauritius took part in the study. There were 28 children from Grade two, 66 children from Grade 4 and 26 children from Grade 6. No one was excluded on the basis of race, religion, or ethnicity. The details on some demographic items with respective percentages and frequencies are presented in Table 1.

Table 1

Frequencies and percentages of demographic items of the participants

Variables

Frequency

Percentage (%)

Gender:

Male

Female

59

61

49.2

50.8

Grades

2

4

6

28

66

26

23.3

55.0

21.7

Religion:

Hindu

Muslim

General Population

27

47

46

22.5

39.2

38.3

Materials and Procedures

After the research was finalised and approved by Kingston University Psychology Ethics Committee, permission to collect data was asked to the Ministry of Education and Scientific Research and the Mauritius Institute of education which are the two bodies in charge of education (includes having control on the schools) in Mauritius. Moreover, prior to the interviews and after the authorisation was obtained from all the headmasters, parental consent forms was sent to the parents and the guardians to ask for their permission to consent their offspring to participate in the study. Seventy two percent of the children returned an approved signed consent form for a total of 120 children. Each one of the later was interviewed individually in a private room by the researcher. (See Appendices for letter to Headmasters and to parents). The interviews were carried out in Mauritian Creole (mother tongue of the island) rather than English because the participants were too young to fully understand the English language. The aims of the study were explained to the participants. They were assured that the interview was confidential and anonymous and that there is no right and wrong answers.

The investigation consisted of three sections:

Demographic data: Basic information such as age, gender, grade, religion and number of siblings, on each child was recorded (See Appendix 1).

Bully-Victim Survey: This interview was conducted to know about the experiences of the participants with their siblings at home and at schools with their peers to be able to determine whether the child is a victim, a bully or a bully-victim. It is an adapted version of Olweus (1991) extensively used questionnaire. The participants were asked whether they had ever been bullied or whether they had bullied others at home or at school in the last six months. Some of the methods were being hit, pushed or kicked, having their belongings snatched from them, being called bad names, being made fun and being excluded from games. Moreover, they had to rate how often they had experienced one or more of the behaviours on a scale of 1 (Never) to 5 (Several times a week) (2: Only once or twice over the 6 months; 3: 2 or 3 times a month; 4: about once a week). They were classified as "pure victim", "pure bully", "bully-victim" or "neutral" (See Appendix 2)

Reasoning of bullying interview: This part of the interview consisted of vignettes about bullying because of gender and ethnicity. In the first part, participants were asked questions with no reference to gender and ethnicity. The second part was based on vignettes concerning experience of children being bullied because of their gender or ethnic group. The researcher has chosen personal interviews over questionnaires mainly because the young children in the sample are not able to fully able to read, understand and write creole and/or English. Note that creole is only a dialect used to converse. Also, the interviewer was able to notice and correct the children's misunderstanding, probed inadequate or vague or incomplete answers despite the length of the interview. Minimum length of the whole interview was 10 minutes and maximum was 40 minutes.

All the questions were designed in English and then translated in Creole (See Appendix 3for English version and Appendix 4 for creole version), which were then back-translated in English by another judge who were blind to the experiment aims and hypotheses. This was to ensure that the meanings were not changed when translated in Creole.

All the questions on bullying were from the three main types of bullying: Direct bullying (includes physical and verbal bullying, Relational bullying and Cyber Bullying. To make the interview easy for the young participants, we first asked the children whether it was alright/acceptable to bully the child in the vignette and then asked by how much by answering 'not a lot' or 'a lot'. Thus, the scores were from a scale of 1 (agree a lot) to 5 (disagree a lot). After that, the children had to justify their answers by giving reasons. All the interviews were audio-recorded and transcribed for further analysis. Common unisex and multi-ethnic Mauritian names were used. Each participant was given a unique code to relate their Demographic questionnaires with their interviews. Children were free to withdraw from the research at any time.

Reliability and Coding

Fifty of the interviews were translated into English and then back to Creole specifically for coding purposes. A different creole speaker from Mauritius made sure the meaning did not change while re-translating. For the last part of the inter As shown in Table 2, the researcher adopted five of Killen et al. (2002) codes of reasoning and added three more categories: stereotypes, reciprocity and Ethics and Pain.

Table 2

Children's reasonings' Justification Categories

Category

Descriptions (and Illustrations of justifications from interviews)

Fairness

Allusion to fairness, equality, and rights (e.g. "It is not fair to do that to a child; we must make friends with everyone; we have no right to fight.")

Empathy

Put emphasis on the feelings and emotions of the child being bullied and/or demonstrating care about the latter. (e.g. "She will feel sad if her brother will not play with her, she will cry.")

Stereotype

Referring to a certain belief that may not be accurate. (e.g. "He must not beat her, she is a girl, and she is weak.")

Group functioning

Situations where every decision is made by the consensus of the whole group or showing concern with how the group will work. (e.g., "The other kids will no longer accept him"; "the other kids will call her the same name.")

Social traditions

Make reference to the norms and expectations of the society and traditions (e.g. "... because a brother is expected to play with his sibling", "we should respect others even if they are different".)

Authority

Bringing up fear of God, Parents, teachers, headmasters, the police or any authority figures. (e.g., "...we must not lie or God will punish us", "he will get in trouble with his mother.")

Reciprocity

Saying that the bully will experience the same thing. (e.g. "The child will retaliate", "the same thing can happen to him.")

Ethics

Making mention of moral codes, principles and values (e.g. " She is still a child", " children should not be beaten", " It is not good to do that"

Pain

Discussing about physical pain, emotional pain or mental pressure. (e.g. " She will get hurt", " he can do bad in exams", " she will stress herself"

With the help of a second judge who good is in both Creole and English and who was blind to the hypotheses, the researcher coded 50 interviews and the remaining 70 interviews were coded separately to obtain inter-rater reliability. An overall kappa of .89 was obtained.

CHAPTER 3

RESULTS and ANALYSIS

The preceding chapter delineates the methodology applied to testing the model. This chapter presents the results of data analysis and research hypotheses testing.

Prevalence of bullying

First analyses show that 22.5% (n=27) of children were classified as neutrals, 28.3% (n=34) as pure victims, 7.5% (n=9) as pure bullies while 41.5% (n=50) as bully-victims. The frequency of each type of bullying is shown in Table 3.

Table 3

Number and relative frequency (%) of pupils involved in frequent Direct, Relational or Cyber- bullying at school.

Overall Bullying

Direct Bullying

Relational Bullying

Cyber Bullying

Neutrals

27(22.5)

47 (39.2%)

45 (37.5%)

112 (93.3 %)

Pure Victims

34(28.3)

34 (28.3%)

32 (26.7%)

6 (5.0%)

Pure Bullies

9(7.5)

16 (13.3%)

9 (7.5%)

1 (.8%)

Bully-Victims

50(41.7)

23 (19.2%)

34 (28.3%)

1 (.9%)

No significant differences in bullying role were found in relation to ethnic groups, gender and grades of participants (p > .05 in each case). However gender was found to be correlated with direct bullying (p<.01)

Types of bullying versus Gender and Ethnicity of Participants. Three separate one-way ANOVA with types of bullying as dependent variable and Gender, Grade and Ethnicity of participant as independent variables were conducted respectively. The first ANOVA revealed a significant difference between Gender and direct bullying, F (1,119) = 7.56, p = .007. More males (M = 1.41, SD = 1.41) were involved in direct bullying than female participants (M = 0.853, SD = 1.05). No significant results were found in the second and third ANOVA with ethnicity and Grade as independent variables

Bullying judgements. In general, mean scores show that children irrespective of grades and ethnicity do not agree with the decision to bully others or being victims of bullying (M > 4.9). Note that the scores are from a scale of 1 (strongly agree) to 5 (strongly disagree) with bullying. On the other hand, the means of scores were compared for bullying against sibling, peer, gender and ethnicity (Table 4)

Table 4

Means of scores for bullying against sibling, peer, gender and ethnicity

Bullying against

Mean (SD)

Ethnicity

4.99 (0.580)

Peer

4.97 (0.119)

Gender

4.94 (0.133)

Sibling

4.91(0.059)

To measure acceptance of judgements for peer and sibling bullying, a 2 (Gender of participant: Male, Female) x 3 (Grade: 2, 4, 6) x 2 (Target of bullying: Sibling, Peer) x 3 (Ethnicity of participant: Hindu, Muslim, General Population) repeated ANOVA was carried out. The results show only ethnicity of participant has a significant interaction with whether the target of bullying is peer and/or sibling, F (2, 69) = 4.44, p = .015. One-way ANOVA revealed a significant effect between ethnicity and bullying against sibling, F (2, 119) = 11.83, p < .001. Post hoc results confirmed the General population group disagree more with sibling bullying (M = 4.97, SD = 4.94) than Hindus (M = 4.72, SD = 0.37)

To amount for the acceptance of judgements for gender and ethnic bullying, a 2 (Gender of participant: Male, Female) x 3 (Grade: 2, 4, 6) x 2 (Target of bullying: Gender, Ethnicity) x 3 (Ethnicity of participant: Hindu, Muslim, General Population) repeated ANOVA was carried out. Neither significant main effects nor significant interactions were found.

Separate one-way ANOVA with Sibling bullying, Peer bullying, Gender bullying and Ethnic bullying as four separate dependent variables and Gender, Grade and Ethnicity of participant as independent variables were conducted respectively. A significant effect was found between Grade and Sibling bullying F (2,119) = 4.89, p = .009. Children from Grade four (M = 4.92, SD = 0.133) and Grade six (M = 4.95, SD = .303) thought that bulling a sibling or being bullied by a sibling was worse than did children in Grade two (M = 4.79, SD = 0.339). Also, another significant result was found between Sibling bullying and bully status. Pure victims (M = 4.97, SD = 0.960) followed by bully-victims (M = 4.91, SD = .200) and Neutrals (M = 4.90, SD = 0.308) found it bad to bully siblings than pure bullies (M = 4.67, SD = 0. 464).

Data Reduction

Children's uses of justifications were collapsed into two groups adopting Killen et al. (2002): Moral (Fairness, Empathy, Reciprocity and Pain) and Social Conventional (Group functioning, Social Traditions, Authority, Stereotypes and Ethics). Stereotype was classified in the social-conventional category as in previous research (Theimer, Killen, & Stangor, 2001). The participants were able to give more than one type of moral or social conventional reasoning. Their answers were coded as 0 if they made mention of it, 1 if they used a reasoning to reject bullying and 1 if they provided a reasoning to support bullying.

Reasoning Judgments. To look at the difference in reasoning, a 2 (Gender of participant: Male, Female) x 3 (Grade: 2, 4, 6) x 2 (Target of bullying: Gender, Ethnicity, Peer, Sibling) x 3 (Ethnicity of participant: Hindu, Muslim, General Population) x 2 (Reason: Moral, Social Conventional) mixed-design ANOVA was carried out. There was a significant main effect of reasoning on the ratings of participants, F (1, 69) = 7.63, p < .005. There was a significant interaction between ethnicity and reasoning, F (2, 103) = 6.40, p < .05. Mauchleys' test indicated that the assumption of sphericity was violated (p < .05). There was a significant Greenhouse-Geiser adjusted interaction between bullying against target (peer, siblings, gender and ethnicity) and reasoning, F (1.85, 119) = 19.04, p < .05; between bullying against targets, reasoning and ethnicity, F (1.85, 119) = 10.53, p < .05 and between bullying, reasoning and Grade, F (1.85, 119) = 2.54, p < .005.

Several one-way ANOVAS of each bully-victim experience with each type of reasoning as per target group were conducted. Experiences were not a significant analyst for children use of reasoning for any situation.

Similarly, one-way ANOVAS were conducted on each bully-victim scenario with each reasonings. For more than three groups, post-hoc results were analysed to see between which two means there were significant results. The results are reported in Table 5.

Table 5:

Bully-victim scenario with each reasoning justification.

Target of bullying

Reasoning

Scenario

F(1,119)

Explanation

Sibling

Fairness

Took things from someone

3.96*

Children who bullied others used more reasoning on fairness to support sibling bullying (M = 1.56, SD = 1.13) than those who did not mention this type of reasoning (M = 1.12. SD = 0.373) For example: " He is my brother, it is okay for me to take his things"

Authority

Children do not want to play/hang out

4.56*

Participants who were victims of exclusion use fear of authority to reject bullying on their siblings more (M= 1.80, SD = 1.16) more than those who did not make mention of it (M= 1.73, SD = 1.07). For example: "She will go and tell mummy, and I will be in trouble."

Reciprocity

Was called bad names

Was tricked in a nasty way

5.66*

9.32**

In both cases, when participants were victims of direct bullying, they feared that there is a higher risk of the same thing happening to the bullies. The respective means of those who used Reciprocity as reasoning are 2.48(1.40) and 1.96(1.06) against 1.82(1.24) and 1.41 (0.755) who did not mention.

Ethics

Hit, Kicked, pushed or threatened someone

Sent upsetting phone calls, photos etc…

6.45*

6.63*

Children who were found to be bullies in these situations did not justified using Ethics reasoning in both scenarios for sibling bullying.

Peer

Empathy

Children do not want to play/hang out

Have received threatening messages etc…

5.73*

4.62*

Participants who used more empathy reasoning (M = 2.44, SD = 1.58 and M = 1.30, SD = 0.776 respectively) than those who did not mention it (M = 1.83, SD = 1.04 and M = 1.07, SD = 0.337 respectively) were found to be victims of those situation. These might reflect their own feelings.

Stereotypes

Took things from someone else

10.56***

The larger number of participants who used stereotypes to support peer bullying (M= 2.50, SD = 2.12) were found to be bullies in this situation. It means that they make use of stereotypes to justify their own actions.

Ethics

Sent upsetting phone calls, etc…

Sent threatening messages, etc…

Cyberbullying

6.74*

5.68*

4.05*

Once again, those who were find bullies of cyber-bullying did not make use of ethics reasoning to reject bullying against peer members.

Gender

Fairness

Relational Bullying

7.02**

Bullies of relational bullying were find to not give reasonings on fairness (M = 1.53, SD = 1.26) than those who mention it (M = 0.940, SD = 1.14)

Empathy

Was called bad names

4.56*

Victims of verbal bullying were once more found to give empathy as a reason why we should not do gender bullying. For example, " She will cry and feel sad if we give her names"

Authority

Was called bad names

Other children told lies

4.19*

7.07**

Victims of those scenarios mentioned that bullies others based on their gender should be afraid of authorities and fear that the same thing can happen to them.

Reciprocity

Was called bad names

5.21*

Ethics

Sent threatening messages, etc…

4.55*

Just like for sibling and peer bullying, more bullies of this type of cyber bullying did not mention ethics as reasoning (M = 1.06, SD = 0.242) than those who did mention (M = 1.00, SD = 0.00)

Pain

Hit, Kicked, pushed or threatened someone

5.18*

Indeed, bullies of this physical bullying did not mention pain (M = 2.49, SD = 0.861) than the other children who used this reasoning to reject gender bullying (M = 2.32, SD = 0.851)

Ethnicity

Fairness

Do not play with someone

14.94***

Those who find it more fair to bully someone from other ethnicity (M = 6.00, SD = 7.07) were found to exclude someone from games than those who did not (M= 1.44, SD = 0.873)

Authority

Was hit, kicked, pushed or threatened by someone

Other children tell lies

4.72*

2.74**

In both cases, victims of these types of bullying mention fear of authority for ethnic bullying.

Stereotypes

Relational bullying

3.26*

Those who gave more support for relational bullying (M = 3.00, SD = 0.00) gave stereotypes as a reasoning to exclude a child based on his/her ethnicity than those who did not make mention of this particular reasoning (M = 1.18, SD = 1.21)

*p < .05 ** p < .01 ***p < .001

Frequencies of reasoning. The table below provides the percentages that the children made use of reasoning in each justification category.

Table 6

Percentages that the children made use of reasoning in each type of reasoning.

Sibling Bullying

% Reject (% support)

Peer Bullying

% Reject (% support)

Gender Bullying

% Reject (% support)

Ethnic Bullying

% Reject (% support)

Total frequency Reject support)

Fairness

37.5(7.5)

35.8(0.8)

45.0

43.3 (1.7)

194 (12)

Empathy

23.3

22.5

25.0

23.3

113

Group Functioning

10 (2.5)

18.3

12.5

14.2

66 (3)

Social Traditions

33.3

22.5

17.5

18.3

110

Authority

25.0

27.5

30.0

31.7

137

Reciprocity

22.5

23.3

20.0

17.5

100

Stereotypes

6.7 (5.8)

7.5 (1.7)

17.5 (14.2)

8.3 (1.7)

48 (28)

Ethics

53.3

64.2

59.2

58.3

282

Pain

33.3

31.7

30.8

32.5

154

Figure 2

Comparison of the percentages of moral and social-conventional reasoning used by children.

CHAPTER 4

DISCUSSION

4.1 Overview

In the following sections, first, we will discuss the findings of the study. Second, limitations of the study will be highlighted and suggestions for future research will be presented.

Irrespective of the grade of the child, his/her ethnic group and gender, no trend was found in the bullying status: Pure bully, pure victim, Bully-victim or neutral. But however it was found that male participants were more involved in direct bullying. This is in line of previous research. Boys in Lagerspetz and Bjorkqvist's (1994) sample, display more of physical aggression than the girls (as mentioned in Mills & Duck, 1999). More recently, in a study to investigate the rate of victimisation and bullying among school students in Turkey, Kartal (2009) found male students were more involved in verbal and physical bullying more than the female students.

From the mean scores, just like the researcher has hypothesised, it was found that most of the students categorically disagree with most types of bullying on other children regardless whether the child is a sibling, peer or someone from the other gender or other ethnicities. This was supported by other studies carried out on children (Tisak, 1995; Killen & Nucci, 1995).

The results also found that there is a significant interaction with ethnicity and bullying against gender and peer. Additional analysis confirmed the difference to be between the General population ethnic group and bullying against siblings. This may be due to the high frequency of broken homes and that children from this ethnic group are most of the times separated from their siblings. Also, it is known that due to their low socio-economic status, the parents have to work for a large number of hours, which result in the siblings taking care of each other. According to a recent report from the Social Security Service (2011), more members from the General Population community are from sub-urbans or villages, live in social housing, are unemployed or are at the bottom end in the labour market. Siblings can be a great source of support in hard family times. Result supported by Dunn et al.(1994a) and Dunn et al. (2001) (Dunn, n.d as mentioned in Smith & Hart, n.d, p. 225-233). There is a lack of research in Mauritius on this matter which can be addressed in the future.

Moreover, it was found that the young participants rated ethnic bullying as more wrong followed by peer bullying, then gender bullying and rated sibling bullying as more acceptable. This is in line with what we foretold. Killen and Stangor (2002) also found that children thought than exclusion based on race is less acceptable than exclusion based on gender. This result was replicated by Killen and colleagues (2002). Along the same line, as children increase in age (grade) it was found they tend to see sibling bullying as less acceptable. These results can be explained by using Dunn's (n.d as mentioned in Smith & Hart, n.d, p. 225-233) reasoning. Small kids think that they have to compromise on their parents' love and that the latter are more concerned with their siblings than themselves which give rise to more sibling conflicts. This reasoning tends to dissolve as the child grows up. Dunn (n.d) also mention that relation with peer embroils more trust, commitment and support that may not exist in all relations between siblings as there is competition in the affection and attention of parents and/or bitterness for being treated differentially. We also found that some participants make more use of stereotypes for Gender bullying. Gender bullying could have been rated as more acceptable than ethnic and peer bullying because of how teachers and parents depend on gender to functionally organise the class or chores at home. The way we ghettoise boys and girls in our school settings can exacerbate gender differences in socialisation which can result in more bullying.

"He is a boy, it is okay for me not to play for him. Actually, my mama herself told me not to play with boys, I will be in trouble if she knows."

Each reasoning category was not found to be a significant analyst for using or not using reasoning for any bully-victim situation. According to Moller and Tenebaum (2011) and Killen et al. (2002), participants are more sensitive to context. So, more analysis was carried out for each of the specific reasoning (such as fairness, stereotypes, empathy) with each scenario from the bully-victim survey. In general, it was found that bullies of specific scenarios are less likely to mention the positive reasoning such as ethics; good group functioning or fear of authority. They use more of fairness to support sibling bullying, ethnic bullying and gender bullying, stereotypes to support peer bullying and ethnic bullying and did make less use of ethics to reject peer bullying and gender bullying. Again, this is in line with what we postulated. We can conclude that bullies of some specific act tend to bully others and use fairness and stereotypes to bully others.

As illustration "She is different and weak, and she cannot play with us because she is too fragile to play wrestling and football; and she does not know how to play counter-strike, does she? So, see, it is alright not to play with her"

On the other hand, victims of bullying made more use of fear of authority and reciprocity to reject sibling, ethnic and gender bullying, more empathy for peer and gender bullying. Victims tend to have more feelings, pity for those being bullied as they have been through the same situations and may know how it feel to be a victim of bullying. Plus, by making more reference to the fear of authority, they insinuate that authority figures are there to give punishment to those that bully others.

Furthermore, children were found to make invoke more of ethics (Social-convention) and fairness and authority (moral convention). More reasoning was used for ethnic followed by gender bullying. Note that in cases where children rate acceptable to bully someone they reason more in terms of stereotypes followed by fairness. Overall, Mauritian children provided more moral reasoning than social-convention reasoning. This is in line to our hypothesis and partly to what Moller and Tenenbaum (2011) found. The latter discovered that Danish children are more likely to use more of than moral reasonings and less social-conventional reasonings in ethnic exclusion but in their sample, to justify gender exclusion, more social-conventional reasoning, like group functioning, was invoked.

No grade differences were found in reasonings. This is opposed to the finding of Killen et al., (2002) where older children were observed to make more reference to multi-faceted reasoning. Similar results were found by Maria (2007). The different types of social reasoning did not vary by grade; more specifically children from 9th grade did not differ from children from 6th grade in the use of social-conventional reasoning. Berk (2006) invoked the Piaget's theory that pre-school and very young children are more likely to reason strictly within moral domain. They have an undeveloped clasp on justice in that they discriminate moral obligations from personal choices and social-conventional matters. From their daily observations and experiences children think that in any scenario, transgressions are wrong (Berk, 2006).

Since the research made reference to ethnic and gender factor, it not impossible for the participants to diminish or hide their prejudice as highlighted Maria (2007) for racially-ethnically situations. Therefore, children might have done so, so as not to appear ethnically biased or biased towards the other gender.

"Everyone has his own religion, we should respect each and every religion"

"We are from an ethnic group because our parents were from the same ethnic group; Sona should not be blamed if she was born from parents with another ethnic group."

Furthermore, how do children act when they are confronted with 'personal priority' and the 'the wrongfulness of bullying'?

From interview: "We should play with everyone, he is a child too, we should not let him alone. (……) Boys and girls should not play together; they play different games, so they should play separately, even if it is same game like 'coukmayé' (hide-and-seek), they should not play together."

This extract expose that how it is possible for a child to give can provide reasoning to disagree with any type of bullying, but still in a different context, the priorities changes. As put forward by the stage theory, reasons do not essential mirror a developmentally "primitive" reaction (Killen, 2007).

The collectivistic Mauritian culture might have had a big influence on our results. Just like Chinese culture, the moral decisions of the Mauritian ch