COSA survey and analysis methods align with the current scientific understanding of best practice in impact assessment but also introduces some comprehensive innovations in the form of its multi-dimensional indicators and data gathering tools. It gathers data directly from farmers, producer organizations, and communities and employs both quantitative and qualitative approaches. COSA applies rigorous practices to determine representative and relatively unbiased samples including control groups. This is also reflected in survey designs that while consistent can also be adapted to particular needs or situations so as to capture local realities while also balancing the requisite quality with reasonable costs. To reduce error and more faithfully capture the realities of human and agricultural systems COSA prefers to assess multiple observations – not just a single snapshot view. COSA’ s approach is well suited to various forms of multi-criteria analysis and it conducts these in collaboration with local and international scientists to ensure the highest level of credibility and usefulness.