This thesis explores the overall question of “What is the influence of Caribbean students’ technology usage, parental background and school type on their self-efficacy?” Different domains of self-efficacy are examined to explore how they might be uniquely affected by technology usage and adolescent student background. This includes entrepreneurial and social self-efficacy, which are underexplored amongst adolescents. Academic self-efficacy is also considered because though it is commonly explored globally, it remains underexplored in the Caribbean and in relation to technology usage. Technology usage (at school for schoolwork, at home for homework and social media time and connections) is explored because of its ubiquity in modern learning contexts, alongside the lack of research on its potential effect on the self-efficacy domains in this study. Parental background (mothers’ education, fathers’ education and wealth), school type and academic achievement are also explored because they have been shown in the global literature to be positively associated with several self-efficacy domains but remain underexplored in the Caribbean context. Exploring technology usage, student background and academic achievement together provides a detailed model of the pathways through which they interact and potentially affect self-efficacy. To explore these issues, a mixed-methods study was adopted, using a cross-sectional questionnaire involving 585 students, as well as semi-structured interviews with 34 teachers, students and a Ministry of Education official. Qualitative classroom and school observations were also done. Data were analysed using Structural equation modelling path analyses and Analyses of variance to establish models of relationships amongst variables. Qualitative data were analysed using thematic analysis to help explain the quantitative results. The results show various significant influences on self-efficacy. In terms of parental background, mothers’ education positively influences all types of self-efficacy. Wealth influences social self-efficacy through increased time spent on social media, and entrepreneurial self-efficacy through increased social media connections. The thematic analyses suggest that a greater exposure to opportunities and ideas beyond their community may be one reason why students with more privileged parental backgrounds might have higher confidence. Technology usage also influences all types of self-efficacy. Academic self-efficacy is directly and positively associated with technology usage for homework, whilst social self-efficacy is positively and directly associated with the time spent on social media and the number of connections on social media. Entrepreneurial self-efficacy is directly and positively associated with the number of connections on social media. Computer use at school had no significant effect on any type of self-efficacy. The thematic analysis suggests that that lack of effect of computer use might be due to insufficient infrastructure and teacher professional development, reducing teachers’ ability to competently facilitate technology use in classrooms. Finally, school ‘prestige’ positively influenced academic self-efficacy. The thematic analysis suggests that ‘prestigious’ schools had greater financial resources and lower levels of bureaucracy in implementing innovative pedagogy, which might facilitate the conditions for academic self-efficacy development. Finally, there were notable differences in the self-efficacy of boys and girls, with boys having higher. The thematic analyses suggest that the socialisation of meekness in girls and boldness in boys may be a reason. Ethnicity results also showed that African students had lower academic self-efficacy than their Indian counterparts, which the thematic analysis suggested could be because African boys are chastised by teachers at a seemingly greater rate. Notably though, African boys had the highest entrepreneurial self-efficacy. The conclusion brings together the results and identifies multiple areas for future research and policy development. Some key conclusions are that teacher professional development and infrastructure development are critical for technology use to be effective for developing student self-efficacy. Mothers’ education also emerged as a key variable, which highlights that mothers should probably be a target for access to further education so that they are better equipped to support their children’s confidence development. Schools could also explore having more celebrations of African culture to help African students feel more academically confident. Finally, the results of this thesis support the usefulness of domain-specific self-efficacy testing, as opposed to the testing of general self-efficacy, since clear differences were found in how each type of self-efficacy is influenced.