A Non-Traditional Randomised Controlled Trial – The IGATE Baseline in Zimbabwe

DFID’s Girls’ Education Challenge prioritises the use of experimental and quasi-experimental evaluations to generate evidence of impact in the learning outcomes and attendance of marginalised girls. The IGATE project is a... [ view full abstract ]

DFID’s Girls’ Education Challenge prioritises the use of experimental and quasi-experimental evaluations to generate evidence of impact in the learning outcomes and attendance of marginalised girls. The IGATE project is a randomised controlled trial being implemented in 467 schools across Zimbabwe, reaching over 90,000 girls. The project’s baseline study invested heavily in the combination of quantitative data with qualitative interviews, a broad gender analysis and an indigenous knowledge study, resulting in an in-depth, multi-faceted interpretation of findings on learning, enrolment, attendance and barriers to girls’ retention in school. While enrolment rates were higher than expected, a careful analysis indicated a pattern of absenteeism, exclusion of older girls, overage, and higher migration rates among girls than boys. EGRA results indicate systematic issues with the use of a second language (English); workload and attendance showed to be key determinants of learning. The triangulation of data sources indicated systematic gaps in quantitative data on violence, marriage and menstrual hygiene. A comparison of quantitative data obtained from caregivers and girls indicates that violence is highly underreported by adults; there is strong indication that early marriage is a hidden issue, masked by high migration rates. The indigenous knowledge study and broad gender analysis provided rich information about perceptions and taboos related to menstruation and virginity in local cultures, supporting the interpretation of data on attendance and retention. This paper will share the results of the baseline study, along with a reflection on the need for consistent application of mixed methods in experimental evaluation designs.