Factors Influencing Female Students’ Intention to Continue Studying Science in Lower Secondary Schools in Dodoma and Mwanza Regions, Tanzania: A Mixed-Methods Study

Background: Gender disparities in science participation may emerge before learners enter upper secondary school; however, evidence on female students’ intentions to continue studying science in Tanzania remains limited. Aim: The study aimed to explore factors associated with female students’ intention to continue studying science in lower secondary schools in the Dodoma and Mwanza regions of Tanzania. Study Design: The study used a convergent parallel mixed-methods design, combining quantitative and qualitative evidence. Place and Duration of Study: It was conducted in six lower secondary schools in the Dodoma and Mwanza regions of Tanzania between January and March 2026. Methodology: A total of 150 female students participated in the study, representing a 100% response rate, and 20 of the 25 invited teachers (80.0% response rate) completed the study. A validated questionnaire with a Likert scale was used to gather quantitative data. Intention to continue studying science was measured using the statement: “When I am given a choice of subjects, I intend to continue studying science subjects,” rated from 1 (strongly disagree) to 5 (strongly agree). The responses were dichotomised as intention to continue (agree/strongly agree) or no intention to continue (neutral/disagree/strongly disagree). Quantitative data were analysed with descriptive statistics and binary logistic regression with school-clustered (robust) standard errors. Semi-structured interviews and structured classroom observations were conducted to collect qualitative data and were then analysed thematically. Results: Overall, 105 students (70.0%) intended to continue studying science and 45 students (30.0%) did not intend to continue studying science. Overall, 68% reported one or more barriers to ongoing participation in science. Teachers' expectations (M = 3.42, SD = 0.61), gendered classroom practices (M = 3.28, SD = 0.55), and perceived assessment bias (M = 3.15, SD = 0.58) had the highest mean barrier scores. In the cluster-adjusted analysis, science self-efficacy (aOR = 2.31, 95% CI [1.48, 3.61]) and family support for STEM (aOR = 2.46, 95% CI [1.51, 4.01]) had positive associations, while science-continuation intention (aOR = 0.43, 95% CI [0.25, 0.74]) had a negative association. Five themes emerged: gendered expectations, assessment practices seen as favouring boys, limited female role models, peer influence, and home environment. Conclusions: Female students’ intention to continue studying science was associated with institutional, pedagogical, social and cultural factors. The results highlight the need for multilevel interventions that tackle teacher attitudes, gender-inclusive classroom and assessment practices, the visibility of female scientists in students’ environments and family engagement to support female students’ sustained engagement in science.