This thesis investigates bullying and negative behaviour among chefs working in commercial kitchens. The idea for this study arose due to evidence from the hospitality industry and amongst chefs in particular which suggested that negative behaviour and bullying were widely accepted practices. However, much of this evidence has been either anecdotal from media reporting or based on small scale studies.The industry has also complained about high labour turnover and the need for a trained workforce. Therefore, this study examines the behaviours to which chefs are exposed and if negative behaviours cause them to leave the industry.This thesis has used a questionnaire to measure responses from chefs who were either in training in catering colleges or working in the industry. Questionnaires were distributed to first year student chefs (n = 202), final year student chefs (n = 153) and working chefs (n = 304). Working chefs and final year student chefs were given questionnaires that included a behavioural method of measuring bullying, known as the NAQ-R, a self labelling method of measuring bullying and items about kitchen specific behaviours. Working chefs and final year student chefs were also asked about positive aspects of work and job satisfaction. First year student chefs were also given a personality instrument as well as being asked their opinions of kitchen specific behaviours. The findings suggest chefs’ exposure to regular bullying was in line with another major UK study (Hoel, 2002). However, occasional bullying was much higher. An examination of industry specific behaviours revealed that chefs tolerate a range of behaviours from verbal abuse on the one hand to physical and sexual abuse on the other. There were also positive aspects about the freedom of work and job satisfaction but this study was unable to find any evidence as to what made chefs stay in the industryThe study found that as student chefs become socialised into their role they were more likely to tolerate negative work behaviours and could identify reasons for their use although, this did not mean that they necessarily agreed with them. In fact, as the working chef sample was very different in ethnicity, gender and nationality from the student sample it could mean that due to negative behaviours on graduation students may not work as chefs.

The ever-increasing use of social media in the internet space have induced a number of problems like cyberbullying and cyberaggression over the internet. Researchers have made a commendable progress on the ongoing fight against cyberbullying but a lot of unresolved issues still persist that primarily motivates the purpose of the research. The paper aims to integrate recent advances in the field of word embedding like fastText, ELMo and stacked flair embeddings combined with a host of robust deep learning techniques to further the efficiency of detection over the state-of-art. Two distinct datasets Formspring and Wikipedia were requested and processed for the purpose of the research. A number of different combinations of word embedding with deep learning methods were tested and compared with CNN with ELMo embedding delivering the most promising results with an F1 score of 0.82 on both datasets. On the other hand, CNN with fastText obtained F1 score of 0.82 on Formspring and 0.64 on Wikipedia dataset but was computationally faster than the counterparts. Moreover, transfer learning was performed using the models to test and prove the robustness and efficacy of the models. The system performed considerably well with superior scores in precision, recall and F1 over the state-of-the-art across all the test cases performed.

Teachers play a critical role in responding to, reporting, and preventing bullying. This study focuses on teachers’ perceptions of various types of bullying: physical, verbal, relational, and especially the more recent phenomenon of cyberbullying. It compares English and US teachers’ perspectives of the seriousness of these different types of bullying, thereby investigating how and why teachers have different views of how to address bullying and the different reasons they give as to why they would or would not respond to different bullying scenarios. This can possibly contribute to potential strategies to combat bullying.

The research is based on a sequential explanatory mixed methods design. In the first quantitative phase, the researcher developed two questionnaires, asking teachers for their perceptions of different hypothetical bullying scenarios, drawing on Expectancy Theory and Social Cognitive Theory. Questionnaires were adapted from those previously developed in a series of American studies (Bauman and Del Rio 2006; Yoon and Kerber 2003; Stauffer et al. 2012). They were administered to a purposive sample of teachers in both England and the US in order to provide a comparative element in two English-speaking country contexts. The questionnaires were analysed to establish whether there are differences between the perceptions and the reasons for responses for new and experienced teachers across the two different country contexts. The analyses also explored teacher perceptions of more long-term strategies to deal with cyberbullying.

In the second qualitative phase, a series of semi-structured interviews were conducted with small sub samples of teachers in both England and the US to provide additional qualitative evidence about teachers’ understandings and experiences of bullying. The interviews explored a number of topics emerging from the quantitative findings pertaining to bullying and cyberbullying, allowing teachers to comment on the survey findings and share their experiences. In addition, the analysis of the qualitative interviews offered the opportunity to provide richer descriptions and insights, building explanations and enhancing understandings. This process enables the research to investigate the degree of similarity between the quantitative and qualitative findings. The combination and integration of findings from both phases of the mixed methods study adds to the ability to triangulate findings and to explore and clarify the reasons for differences in perspectives between English and US teachers. In the mixed sequential explanatory strategy, quantitative findings reveal general trends from the data that are later expanded and investigated further within the qualitative phase. The discussion then explores the nature of the similarities and differences between these findings.

The quantitative findings suggest that three main variables predicted both US and English teachers’ responses to cyberbullying scenarios: the perceived seriousness of the situation, the confidence in one’s ability, and the location (home having less of a response than school scenarios). The qualitative findings suggest that teachers’ perceptions of severity could be shaped by current policies and the relationships teachers developed with their students; the perceived confidence in one’s self could be mediated by one’s interpretation of the school culture and the trust established with others on staff; responding to a home cyberbullying situation could depend on one’s understanding of pastoral responsibilities and the current parental involvement. This research provides new findings that can inform policy makers and practitioners in designing and implementing anti-bullying programs that are relevant to different contexts. Knowledge of teacher perceptions could prove useful as a starting point for developing a teacher education program. An understanding of the different types of bullying and cyberbullying that need attention could enable anti-bullying programs to present new and relevant material.

The overarching aim of the thesis was to identify and explore a behavioural typology of the use of reactive and proactive aggression in a sample of 9-14 year-old English children and adolescents. To date, few studies have employed a person centred approach to investigate behavioural patterns of the use of both reactive and proactive aggression. Of these only two have investigated the behaviour of community, rather than specialised or clinical participant samples (Crapanzano, Frick and Terranova 2010; Mayberry and Espelage 2007). However, these two studies employed methods which raise questions regarding the reliability and/or generalisability of the results obtained. For example, neither study asked participants whether they had actually engaged in the behaviours of interest; rather they asked children to report on how likely they felt they were to react in the same way as described in a list of aggressive scenarios presented to them. As such the studies did not actually record engagement in aggressive behaviour, rather the participants’ perceived likelihood that they would behave in a certain way. Furthermore, neither study was conducted in the UK, leading to questions of generalisability between participant samples. Both research and school policy in England and Wales has focused on exploring the use of proactive forms of aggression (including bullying) in schools, and reactive aggression has to date been neglected. However, it is essential that we identify the prevalence and patterns of the use of both reactive and proactive forms of aggression as both are prevalent in schools and place children and adolescents at risk of harm. Employing a mixed methodological approach, a two-phased data collection procedure was followed to identify and explore a behavioural typology of the use of reactive and proactive aggression and differences in associated demographic, behavioural and socio-cognitive risk factors between the behavioural groups identified. In Phase 1, focus groups were conducted with 57 (20 males, 37 females) children and adolescents aged 9 – 18 years, in order to understand how they define terminology utilised across the research literature to describe acts of negative interpersonal behaviour. Across three data collection sites participants reported consistent definitions of the terms provided to them and differentiated between the terms aggression, violence and bullying. Social representations of the reasons they believed people engaged and avoided engaging in interpersonal aggression also emerged from their talk. These related to the role of taking the perspective, or empathising with others and the perception of a level of justification for certain types of behaviour, enacted under certain conditions. In Phase 2 a survey design was used to collect both qualitative and quantitative data from 658 children and adolescents aged 9-14 years (302 males, 356 females). The aim of Phase 2 was to identify a behavioural typology of the use of reactive and proactive aggression based on self-report data collected using a modified version of the Reactive-Proactive Aggression Questionnaire (RPAQ; Raine et al 2006). Once the behavioural subtypes were identified, associations between subtype and, involvement in bullying relationships, demographic (age and gender) and social-cognitive characteristics (empathy, perceived acceptance of behaviour and social representations of why people become involved in negative interpersonal interactions) were examined. Cluster analysis of the RPAQ data identified three distinct behavioural groups characterised by lower than the sample median use of both types of behaviour (Low Aggression: characterising 57.1% of the sample), Moderate-high reactive and Low-moderate proactive aggression (characterising 34.4% of the sample), and finally a group indicating frequent use of both reactive and proactive aggression (High Aggression: characterising 8.5% of the sample). The only age and gender related differences within the clusters were found in the low frequency aggression cluster. Specifically, there were a greater proportion of females compared to males in this cluster. The only age related difference found was a greater percentage of primary school children compared to 13-14 year olds in the Low aggression cluster. Group membership was found to be associated with self-reported bullying as measured by the Peer Victimization and Bullying Scale (Mynard and Joseph 2000). The High frequency aggression cluster contained a significantly higher percentage of those indicating being a bully or a bully-victim compared to the other two clusters. Whereas the Low frequency cluster contained a significantly higher percentage of those indicating not being involved in bullying compared to the other two clusters. However, reporting being a victim of bullying was not associated with any one of the three clusters. Of the socio-cognitive variables, a significant incremental increase was found in the perceived acceptance of both reactive and proactive aggression as the reported frequency of the use of both types of behaviour increased across the three behavioural groups (as measured by a modified version of the RPAQ; Raine et al 2006). Conversely, an incremental decrease was observed between the frequency of the use of aggression and reported affective empathy (as measured by the Basic Empathy Scale; Jolliffe and Farrington 2006), with a significant difference being found between the Low and High frequency aggression groups. No significant differences between the groups in self-reported cognitive empathy were found. Finally, participants were asked two open-ended questions relating to their perception of why people are ‘picked on’, or ‘pick on’ others. Thematic Analysis identified a number of social representations held across the participant sample, with a further content analysis identifying that there were no significant differences in the extent to which these representations were endorsed by the three behavioural groups. The findings of the current research have important implications for our understanding of the developmental pathways for the use of reactive and proactive aggression. They identify that both types of behaviour co-occur, suggesting that the risk factors for the development of these two types of behaviour may not be so distinct and/or the risk factors associated with each are likely to co-occur. Consequently, school behaviour policies need to include strategies for addressing both forms of aggression. Interventions to reduce/prevent this behaviour need to be designed to address the risk factors which are promoting the specific motivations of both reactive and proactive behaviour.