Ravens Consulting was sub-contracted by Sokoine University of Agriculture for the electronic authoring of the ATONU survey. Their responsibilities included monitoring data collection, setting up servers, managing data, and overseeing the quality of data collection and analysis.
The ATONU project, which stands for Agriculture to Nutrition, is an African initiative focused on improving nutrition outcomes for smallholder farm families through robust, evidence-based interventions. This initiative was implemented in three countries: Ethiopia, Nigeria, and Tanzania.
The data analysis in the ATONU project involved computing various outcome variables, such as wealth index, dietary diversity for households, women, and children, women empowerment indicators, nutritional indicators, food security, food shortages, WASH practices, vegetable consumption, and general nutrition knowledge. These computations were crucial for assessing the impact of the interventions.
To compare the changes from baseline to endline across different treatment arms, a Difference in Difference (D-I-D) analysis was conducted. This analysis used both generalized estimating equations and linear mixed models for binary and continuous outcomes, respectively. The models took into account the cluster randomization of villages and the repeated measures of outcome variables.
Fixed effects in the models included variables such as zones, improved chicken ownership, and treatment arms, while village was treated as a random effect. The statistical analysis was performed using the Statistical Analysis System (SAS) software version 9.4, ensuring a rigorous examination of the data and the effectiveness of the interventions.