Threshold Performance Evaluation 2009-2010

The methodologies employed by the evaluation are both quantitative and qualitative. Wherever possible, we obtain data and employ an identification strategy with which to estimate the causal effect of a given intervention. Data are collected from program participants, and data collection was handled by EcomResearch – a Dar es Salaam based research firm. Because data is often unavailable, and because randomization was never employed in treatment assignment, such causal estimations are sometimes impossible. Where quantitative methods cannot be used to assess the effect of an intervention, we employ qualitative methods. We conduct interviews with program participants, engage in document review, and – in some cases – conduct semi-structured field interviews to assess the implementation and sustainability of a given intervention. We also examine the logical link between the intervention and the reduction of corruption, in light of related academic literatures. The quantitative methods employed in this evaluation give precise estimates of the effect of training on a narrowly defined quantity of interest. Qualitative methods, on the other hand, cannot give as precise an estimate, but may incorporate a broader variety of factors than quantitative estimation allows.