Generative AI as a Collaborative Partner Empowering Tanzanian Secondary School Teachers to Deliver Personalised Learning and Overcome Barriers to Equitable Adoption

Abstract

Many educators face overcrowded classrooms and limited resources, making it challenging to meet diverse student needs. This study examines how generative AI can assist secondary school teachers in Tanzania to create personalised learning paths more efficiently and effectively. Employing a convergent parallel mixed-methods design, the study integrated concurrent quantitative experimental elements (pre- and post-intervention assessments of student engagement, performance, and teacher workload via paired t-tests) with qualitative data (teacher feedback, usage logs, open-ended surveys, and focus groups). To ground findings in authentic contexts, the qualitative strand drew on bounded, site-specific insights from participating schools in Dar es Salaam (urban) and Dodoma (semi-urban/rural elements). Data were collected from 120 teachers who tested accessible AI tools such as ChatGPT and Grok for lesson planning, assessment, and adaptive content delivery. Quantitative data underwent paired t-tests; qualitative data were analysed through thematic coding; integration occurred at interpretation for triangulation. Results showed significant improvements in student engagement (from 2.8 to 4.3 on a 5-point scale) and academic performance (test scores increased from 61% to 75%), alongside a reduction in teacher workload. Teachers perceived the tools as intuitive and beneficial for customising instruction, though challenges such as inadequate training and infrastructure (especially in rural areas) persisted. The study concludes that generative AI offers a scalable, inclusive solution when supported by professional development and digital upgrades. It recommends strategic investments in AI literacy training, rural connectivity, and localised ethical guidelines to realise its potential and address equity gaps in Tanzanian schools.