Background: Understanding student attitudes to curriculum subjects is central to providing evidence-based options to policy makers in education. Purpose: We illustrate how quantitative approaches used in the social sciences and based on multivariate analysis (categorical Principal Components Analysis, Clustering Analysis and General Linear Modelling), can complement qualitative analysis to support this need. Sample: Our example involved an attitude survey of 128 students from five high schools across Botswana to Design and Technology (DT), this subject having declined in uptake over 10 years by up to 6% per year, despite positive encouragement by the government. Design and methods: Qualitative interviews, carried out concomitantly, indicated consistently that age, gender and school performance all affected attitudes. Multivariate analysis provided additional information in ranking how different attitudes contribute to the overall perception of a subject (PCA-Factor analysis), in assessing the relative and interacting effects of external determinants like age or gender and classifying students into attitude groups. Results: Our findings show that DT enrolment could be improved by targeting children who deemed DT too difficult or unimportant as a subject, and by reinforcing perceptions of DT as an enjoyable life-skill. Conclusions: We suggest that combined quantitative and qualitative analysis can act as an effective, evidence-based means to inform educational policies. (Contains 4 figures and 4 tables.)